BJBehnam Jafarpour

Behnam Jafarpour

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Behnam Jafarpour has a total of 57 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Deep Neural Networks for Fast Prediction of Flow and Geomechanical Effects during Geologic CO2 Storage

    by Fangning Zheng, Birendra Jha, Behnam Jafarpour

    Summary: Numerical simulation is commonly used to predict the response of geologic storage formations to carbon dioxide (CO2) injection, including rock deformation and CO2 plume migration. It provides a useful tool for site screening and implementing standard field development workflows, such as well location optimization. However, coupled physics models are computationally demanding, rendering them impractical for tasks that involve thousands of geologic and development scenarios. We study the feasibility of using deep learning models to significantly reduce the computational overhead associated with simulating and quantifying the geomechanical risks of CO2 storage (i.e., ground surface displacement and subsurface rock failure). Using simulated data, we train a U-Net convolutional neural network to learn a mapping from well locations and spatially distributed model parameters to the simulation outputs of interest, including plastic strain and CO2 saturation that exhibit strong nonlinearity and sharp gradients in space-time. Our proposed architecture maps different well locations to the corresponding pressure fields, CO2 saturation, vertical displacement, and plastic strain. The U-Net model can be adopted as an efficient tool for the site screening and identification of injection well locations to minimize geomechanical risks. The results show that the U-Net model can dramatically reduce the computational cost of well placement workflows by replacing coupled physics simulation with a fast proxy model that can be used to predict the geomechanical risk associated with different well locations and injection strategies. The developed framework can be used to reduce the computational cost of coupled-physics modeling and facilitate its application to decision-making workflows and field management.


  • Transformer Neural Networks for Behavior-Centric Production Forecasting in Unconventional Reservoir

    by Jodel Cornelio, Syamil Mohd Razak, Young Cho, Hui-Hai Liu, Ravimadhav Vaidya, Behnam Jafarpour

    Summary: Data-driven models, such as neural networks, provide an alternative to physics-based simulations in predicting well behavior within unconventional reservoirs. However, these models struggle to interpret the factors behind production performance, the underlying physics of the process, and correlations unique to individual wells and their flow behaviors. These models are designed to approximate general trends present in data sets, which can hinder their ability to learn behaviors specific to certain data segments. In large unconventional fields, wells within a single formation can exhibit varied production behaviors, emphasizing the need to segment the data into more relevant subsets. Such data segmentation enables the development of local models (LMs) that capture regional or behavioral correlations. In contrast, a field-wide (global) AI model, trained on the entirety of the data, tends to produce averaged predictions, which can dilute specific correlations and reduce the overall predictive power. This work uses a transformer neural network to enable a behavior-based clustering process in predicting production from unconventional fields. By leveraging the attention mechanism inherent in transformers, we extract attention scores that link well properties to production behaviors. The information from these attention scores improves the clustering quality, which is crucial for training data-driven models on segmented data sets. The proposed approach highlights the advantages of developing behavior-centric (local) models that provide higher accuracy by capturing correlations distinct to a particular region or production behavior. Notably, the attention mechanism facilitates the isolation and curation of data subsets for training behavior-centric models, thereby improving prediction performance.


  • Open Access

    A Fluid Flow‐Based Deep Learning (FFDL) Architecture for Subsurface Flow Systems With Application to Geologic CO 2 Storage

    by Zhen Qin, Yingxiang Liu, Fangning Zheng, Behnam Jafarpour

    Prediction of the spatial‐temporal dynamics of the fluid flow in complex subsurface systems, such as geologic storage, is typically performed using advanced numerical simulation methods that solve the underlying governing physical equations. However, numerical simulation is computationally demanding and can limit the implementation of standard field management workflows, such as model calibration and optimization. Standard deep learning models, such as RUNET, have recently been proposed to alleviate the computational burden of physics‐based simulation models. Despite their powerful learning capabilities and computational appeal, deep learning models have important limitations, including lack of interpretability, extensive data needs, weak extrapolation capacity, and physical inconsistency that can affect their adoption in practical applications. We develop a Fluid Flow‐based Deep Learning (FFDL) architecture for spatial‐temporal prediction of important state variables in subsurface flow systems. The new architecture consists of a physics‐based encoder to construct physically meaningful latent variables, and a residual‐based processor to predict the evolution of the state variables. It uses physical operators that serve as nonlinear activation functions and imposes the general structure of the fluid flow equations to facilitate its training with data pertaining to the specific subsurface flow application of interest. A comprehensive investigation of FFDL, based on a field‐scale geologic storage model, is used to demonstrate the superior performance of FFDL compared to RUNET as a standard deep learning model. The results show that FFDL outperforms RUNET in terms of prediction accuracy, extrapolation power, and training data needs.


  • Open Access

    Improving the Parameterization of Complex Subsurface Flow Properties With Style‐Based Generative Adversarial Network (StyleGAN)

    by Wei Ling, Behnam Jafarpour

    Representing and preserving complex (non‐Gaussian) spatial patterns in aquifer flow properties during model calibration are challenging. Conventional parameterization methods that rely on linear/Gaussian assumptions are not suitable for representation of property maps with more complex spatial patterns. Deep learning techniques, such as Variational Autoencoders (VAE) and Generative Adversarial Networks (GAN), have recently been proposed to address this difficulty by learning complex spatial patterns from prior training images and synthesizing similar realizations using low‐dimensional latent variables with Gaussian distributions. The resulting Gaussian latent variables lend themselves to calibration with the ensemble Kalman filter‐based updating schemes that are suitable for parameters with Gaussian distribution. Despite their superior performance in generating complex spatial patterns, these generative models may not provide desirable properties that are needed for parameterization of model calibration problems, including robustness, smoothness in the latent domain, and reconstruction fidelity. This paper introduces the second generation of style‐based Generative Adversarial Networks (StyleGAN) for parameterization of complex subsurface flow properties and compares its model calibration properties and performance with those of the convolutional VAE and GAN architectures. Numerical experiments involving model calibration with the Ensemble Smoother with Multiple Data Assimilation (ES‐MDA) in single‐phase and two‐phase fluid flow examples are used to assess the capabilities and limitations of these methods. The results show that parameterization with StyleGANs provides superior performance in terms of reconstruction fidelity and flexibility, underscoring its potential for improving the representation and reconstruction of complex spatial patterns in subsurface flow model calibration problems.


  • Dynamic Physics-Guided Deep Learning for Long-Term Production Forecasting in Unconventional Reservoirs

    by Syamil Mohd Razak, Jodel Cornelio, Young Cho, Hui-Hai Liu, Ravimadhav Vaidya, Behnam Jafarpour

    Summary: Neural network predictive models are popular for production forecasting in unconventional reservoirs due to their ability to learn complex relationships between well properties and production responses from extensive field data. The intricate flow behavior in hydraulically fractured unconventional reservoirs, which remains poorly understood, makes these statistical models particularly useful. Various neural network variants have been developed for production prediction in these reservoirs, each offering predictive capability of varying levels of granularity, accuracy, and robustness against noisy and incomplete data. Neural network predictive models that integrate physical principles are especially useful for subsurface systems, as they provide predictions that adhere to physical laws. This work introduces a new dynamic physics-guided deep learning (DPGDL) model that incorporates physical functions into neural networks and employs residual learning to compensate for the imperfect description of the physics, under variable data support. The new formulation allows for dynamic residual correction, avoids unintended bias due to less-than-ideal input data, and provides robust long-term predictions. The DPGDL model improves upon a static formulation by utilizing a masked loss function to enable learning from wells with varying production lengths and by improving the results when partially-observed timesteps are present. In addition, a sequence-to-sequence residual model has been developed to correct additional biases in the long-term predictions from the physics-constrained neural networks. Several synthetic data sets with increasing complexity as well as a field data set from the Bakken are used to demonstrate the performance of the new DPGDL model.


  • Open Access

    Fluid Flow-based Deep Learning (FFDL) for geologic CO 2 Storage

    by Zhen Qin, Yingxiang Liu, Fangning Zheng, Behnam Jafarpour

    Carbon capture and storage (CCS) is one of the few strategies for reducing CO 2 emissions by injecting it into deep geologic formations. The injection of CO 2 into heterogeneous rock formations triggers complex coupled flow and transport processes that are not trivial to describe and predict. Advanced numerical simulation is often used as a standard tool to predict the spatial-temporal evolution of CO 2 plume and induced pressure changes. However, numerical simulation is computationally demanding, limits the use of standard field management workflows, and hinders real-time analysis and decision-making for risk mitigation. Standard deep learning models provide powerful alternative prediction tools. However, they have important limitations, including lack of interpretability, extensive data needs, and physical inconsistency. To overcome these limitations, a Fluid Flow-based Deep Learning (FFDL) architecture is presented for spatial-temporal prediction of the injected CO 2 in storage formations. The architecture of FFDL consists of a physics-based encoder to construct physically meaningful latent variables, and a residual-based processor to predict the evolution of state variables. The FFDL model uses physical operators that serve as nonlinear activation functions and impose hard constraints to respect the general structure of the fluid flow equations. A comprehensive investigation of FFDL, based on a field-scale saline aquifer, is used to demonstrate its superior performance compared to standard deep learning models. The results show that FFDL exhibits strong generalization capability and provides more reliable and physically consistent predictions of CO 2 plume migration. The flexibility of FFDL makes it suitable for various applications, including decision-making, optimization, and inverse modeling.

  • Physics-Guided Deep Learning for Improved Production Forecasting in Unconventional Reservoirs

    by Syamil Mohd Razak, Jodel Cornelio, Young Cho, Hui-Hai Liu, Ravimadhav Vaidya, Behnam Jafarpour

    Summary: The complexity of physics-based modeling of fluid flow in hydraulically fractured unconventional reservoirs, together with the abundant data from repeated factory-style drilling and completion of these resources, has prompted the development and application of data-driven statistical models for predicting hydrocarbon production performance. More recently, machine learning algorithms have been widely studied in developing data-driven prediction models for unconventional reservoirs. These models often require a large amount of high-quality training data with sufficient range to avoid excessive extrapolation and produce reliable predictions. Unlike statistical models, physics-based models represent causal relations between input and output variables to provide predictions beyond available data. While a detailed physics-based description of fluid flow in unconventional reservoirs is not yet available, approximate physical flow functions have been proposed to capture the general production behavior of unconventional wells. These physical functions can be augmented with the available data to improve data-driven methods by constraining the models to adhere to the general production trends. In this paper, we develop a physics-constrained data-driven model by embedding physical flow functions into neural network models. Since the performance of the physics-constrained model depends on the relevance of the embedded physics, a sizeable residual prediction error is expected if the collected data do not sufficiently match the embedded model. The residual model typically represents errors in the description of inputs or any missing physical phenomenon. We compensate for such errors through residual learning, where an auxiliary neural network is designed to learn the complex relationship between the input parameters (such as formation and completion properties) and the expected prediction residuals. The new physics-guided deep learning (PGDL) model augments any physics-constrained prediction model with residual learning to increase its prediction accuracy. Several synthetic and actual field data sets, from the Bakken play, are used to demonstrate the performance of the PGDL model.


  • Transfer Learning with Prior Data-Driven Models from Multiple Unconventional Fields

    by Jodel Cornelio, Syamil Mohd Razak, Young Cho, Hui-Hai Liu, Ravimadhav Vaidya, Behnam Jafarpour

    Summary: Constructing reliable data-driven models to predict well production performance (e.g., estimated ultimate recovery, cumulative production, production curves, etc.) for unconventional reservoirs requires large amounts of data. However, when considering unconventional reservoirs in their early stages of development, where data and the wells drilled are limited, one may benefit from leveraging available data and/or pretrained models from other more developed fields. Transfer learning, the process of storing knowledge gained while solving one problem (source data) and applying it to solve a different but related problem (target data), provides a workflow for alleviating data needs in training a data-driven model in fields with limited data. However, a pitfall in the application of transfer learning is the possibility of negative transfer, that is, transferring incorrect or irrelevant knowledge to the target data. In particular, the black-box nature of most data-driven models, e.g., neural networks, support vector machines, and random forest, makes it difficult to completely interpret the contribution of different source models used for knowledge transfer. Hence, ranking the viability of source models for transfer learning can reduce the risk of negative transfer and improve the prediction performance. In this paper, we illustrate the impact of negative transfer and how it can be identified, and present a new approach for ranking multiple source models based on their positive transfer contribution. Finally, we propose a framework to build a reliable model to predict well production performance by combining multiple sources of information into one network to be transferred and retrained with limited data in fields at their early stages of development.