View in EDS

WellPINN: Accurate Well Representation for Transient Fluid Pressure Diffusion in Subsurface Reservoirs With Physics‐Informed Neural Networks.

Saved in:
Bibliographic Details
Title: WellPINN: Accurate Well Representation for Transient Fluid Pressure Diffusion in Subsurface Reservoirs With Physics‐Informed Neural Networks.
Authors: Walter, Linus1,2,3 (AUTHOR) linus.walter@csic.es, Kong, Qingkai4 (AUTHOR), Hanson‐Hedgecock, Sara1 (AUTHOR), Vilarrasa, Víctor1 (AUTHOR) victor.vilarrasa@csic.es
Source: Water Resources Research. May2026, Vol. 62 Issue 5, p1-15. 15p.
Subjects: Domain decomposition methods, Fluid flow, System identification, Petroleum engineering, Unsteady flow
Abstract: Accurate representation of pumping wells is essential for reliable reservoir characterization and simulation of operational scenarios in subsurface flow models. Physics‐informed neural networks (PINNs) are emerging as a promising alternative to numerical models for reservoir modeling, offering seamless integration of monitoring data and governing physical equations. However, existing PINN‐based studies face major challenges in capturing fluid pressure near wells when using a source/sink term, particularly during the early stages after pumping begins. We address this problem by introducing WellPINN, a workflow in which an initially trained PINN infers fluid pressure across the entire reservoir domain using a large equivalent well radius. This initial PINN solution is then locally refined around the well by a set of subdomain PINNs that are trained for smaller equivalent well radii. Continuity across these subdomain interfaces as well as at the initial condition is ensured by hard‐constraining each PINN on its subdomain boundary. Our results demonstrate WellPINN as the first workflow of its kind to focus on accurate inference of fluid pressure from pumping rates throughout the entire injection period, significantly advancing the potential of PINNs for inverse modeling and operational scenario simulations. All data and code for this paper are openly available at https://doi.org/10.20350/DIGITALCSIC/17260. Plain Language Summary: Accurately representing wells is crucial when building reservoir models to simulate fluid flow. During well testing, it is particularly important to match applied flow rates and observed pressures to calibrate the model. A promising tool are physics‐informed neural networks (PINNs) that flexibly combine measured data with physical laws. However, existing PINNs often fail to capture the fluid pressure field near wells, especially immediately after injection begins. To overcome this limitation, we developed a new workflow called WellPINN. It uses several PINN models in sequence. The first model is trained for the whole modeling domain, while each subsequent model focuses on a smaller area around the well. Simultaneously, we also reduce the dimensions of the representative well. We demonstrate WellPINN on a test case involving a single pumping well in a two‐dimensional reservoir. Our results show that using three PINNs in sequence can successfully model pressure changes from the reservoir boundary to the well center across multiple spatial scales. WellPINN is the first PINN‐based method to accurately predict pressure throughout the entire injection period. This opens new possibilities for applying PINNs in inverse modeling and planning real‐world reservoir operations, including in more complex reservoir settings or with multiple wells. Key Points: WellPINN is a sequential training workflow that uses physics‐informed neural networks (PINNs) to solve transient fluid flow problemsWe accurately represent a well at a reservoir‐scale domain that spans three orders of magnitude in space dimensionDomain decomposition and logarithmic time scaling are required to accurately represent pumping wells in PINNs with tanh activation functions [ABSTRACT FROM AUTHOR]
Copyright of Water Resources Research is the property of Wiley-Blackwell and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)
Database: Engineering Source
Description
Abstract:Accurate representation of pumping wells is essential for reliable reservoir characterization and simulation of operational scenarios in subsurface flow models. Physics‐informed neural networks (PINNs) are emerging as a promising alternative to numerical models for reservoir modeling, offering seamless integration of monitoring data and governing physical equations. However, existing PINN‐based studies face major challenges in capturing fluid pressure near wells when using a source/sink term, particularly during the early stages after pumping begins. We address this problem by introducing WellPINN, a workflow in which an initially trained PINN infers fluid pressure across the entire reservoir domain using a large equivalent well radius. This initial PINN solution is then locally refined around the well by a set of subdomain PINNs that are trained for smaller equivalent well radii. Continuity across these subdomain interfaces as well as at the initial condition is ensured by hard‐constraining each PINN on its subdomain boundary. Our results demonstrate WellPINN as the first workflow of its kind to focus on accurate inference of fluid pressure from pumping rates throughout the entire injection period, significantly advancing the potential of PINNs for inverse modeling and operational scenario simulations. All data and code for this paper are openly available at https://doi.org/10.20350/DIGITALCSIC/17260. Plain Language Summary: Accurately representing wells is crucial when building reservoir models to simulate fluid flow. During well testing, it is particularly important to match applied flow rates and observed pressures to calibrate the model. A promising tool are physics‐informed neural networks (PINNs) that flexibly combine measured data with physical laws. However, existing PINNs often fail to capture the fluid pressure field near wells, especially immediately after injection begins. To overcome this limitation, we developed a new workflow called WellPINN. It uses several PINN models in sequence. The first model is trained for the whole modeling domain, while each subsequent model focuses on a smaller area around the well. Simultaneously, we also reduce the dimensions of the representative well. We demonstrate WellPINN on a test case involving a single pumping well in a two‐dimensional reservoir. Our results show that using three PINNs in sequence can successfully model pressure changes from the reservoir boundary to the well center across multiple spatial scales. WellPINN is the first PINN‐based method to accurately predict pressure throughout the entire injection period. This opens new possibilities for applying PINNs in inverse modeling and planning real‐world reservoir operations, including in more complex reservoir settings or with multiple wells. Key Points: WellPINN is a sequential training workflow that uses physics‐informed neural networks (PINNs) to solve transient fluid flow problemsWe accurately represent a well at a reservoir‐scale domain that spans three orders of magnitude in space dimensionDomain decomposition and logarithmic time scaling are required to accurately represent pumping wells in PINNs with tanh activation functions [ABSTRACT FROM AUTHOR]
ISSN:00431397
DOI:10.1029/2025WR041404