Using Karhunen-Loeve decomposition and artificial neural network to model miscible fluid displacement in porous media

In this paper, we describe an approach to model fluid displacements in poro us media that combines two powerful techniques, namely Karhunen-Loeve (KL) decomposition and artificial neural networks (ANNs). The KL, decomposition, Tor data compression and feature identification, is used to extract cohere nt structures or eigenfunctions using fluid concentration maps obtained fro m fine-mesh numerical simulations of miscible fluid displacements of oil by solvent in a two-dimensional vertical cross-section. Twenty KL eigenfuncti ons that capture 98.8% of the total energy are extracted. Corresponding dat a coefficients are constructed by projecting the fluid concentration maps o f the numerical simulations onto the KL eigenfunctions. Processing these da ta coefficients through an ANN is found to be a powerful tool in predicting the fluid displacements of the fine-mesh numerical simulations without act ually performing these simulations. (C) 2000 Elsevier Science Inc. All righ ts reserved.