Neural networks and their derivatives for history matching and reservoir optimization problems

Bruyelle, Jérémie; Guérillot, Dominique

doi:10.1007/s10596-013-9390-y

Cited by 12 publications

(3 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…An artificial neural network is a mathematical model composed of several computing units (neuron) connected to each other [3,4]. Each neuron is an elementary process that receives a number of inputs from other neurons.…”

Section: Artificial Neural Network Designmentioning

confidence: 99%

Geochemical equilibrium determination using an artificial neural network in compositional reservoir flow simulation

Guérillot

Bruyelle

2019

Comput Geosci

Self Cite

View full text Add to dashboard Cite

The fluid injection in sedimentary formations may generate geochemical interactions between the fluids and the rock minerals, e.g., CO 2 storage in a depleted reservoir or a saline aquifer. To simulate such reactive transfer processes, geochemical equations (equilibrium and kinetics equations) are coupled with compositional flows in porous media in order to represent, for example, precipitation/dissolution phenomena. The aim of the decoupled approach proposed consists in replacing the geochemical equilibrium solver with a substitute method to bypass the huge consuming time required to balance the geochemical system while keeping an accurate equilibrium calculation. This paper focuses on the use of artificial neural networks (ANN) to determine the geochemical equilibrium instead of solving geochemical equations system. To illustrate the proposed workflow, a 3D case study of CO 2 storage in geological formation is presented.

show abstract

Section: Artificial Neural Network Designmentioning

confidence: 99%

Geochemical equilibrium determination using an artificial neural network in compositional reservoir flow simulation

Guérillot

Bruyelle

2019

Comput Geosci

Self Cite

View full text Add to dashboard Cite

show abstract

“…to reservoir engineering and fluid dynamics [18], [19], [20]. Recently, Bruyelle et al (2014) [21] applied the neural network-based RBF to obtain the first-order and second-order derivative information of a reservoir model and estimate the gradients and Hessian matrix for reservoir production optimization. The accuracy of RBF-based gradient approximation is determined by the sampling strategies of the interpolation data [21].…”

Section: Introductionmentioning

confidence: 99%

“…Recently, Bruyelle et al (2014) [21] applied the neural network-based RBF to obtain the first-order and second-order derivative information of a reservoir model and estimate the gradients and Hessian matrix for reservoir production optimization. The accuracy of RBF-based gradient approximation is determined by the sampling strategies of the interpolation data [21]. For high dimensional problems, the classical global RBF interpolation algorithm requires a large num-ber of interpolation data to capture the flow dynamic as much as possible [22].…”

Section: Introductionmentioning

confidence: 99%

Non-intrusive subdomain POD-TPWL for reservoir history matching

et al. 2018

View full text Add to dashboard Cite

This paper presents a non-intrusive subdomain POD-TPWL (SD POD-TPWL) algorithm for reservoir data assimilation through integrating domain decomposition (DD), radial basis function (RBF) interpolation and the trajectory piecewise linearization (TPWL). It is an efficient approach for model reduction and linearization of general non-linear timedependent dynamical systems without intruding the legacy source code. In the subdomain POD-TPWL algorithm, firstly, a sequence of snapshots over the entire computational domain are saved and then partitioned into subdomains. From the local sequence of snapshots over each subdomain, a number of local basis vectors is formed using POD, and then the RBF interpolation is used to estimate the derivative matrices for each subdomain. Finally, those derivative matrices are substituted into a POD-TPWL algorithm to form a reduced-order linear model in each subdomain. This reduced-order linear model makes the implementation of the adjoint easy and resulting in an efficient adjoint-based parameter estimation procedure. The performance of the new adjoint-based parameter estimation algorithm has been assessed through several synthetic cases. Comparisons with the classic finite-difference based history matching show that our proposed subdomain POD-TPWL approach is obtaining comparable results. The number of full-order model simulations required is roughly 2-3 times the number of uncertain parameters. Using different background parameter realizations, our approach efficiently generates an ensemble of calibrated models without additional full-order model simulations.

show abstract