Artificial Neural Network to Forecast Enhanced Oil Recovery Using Hydrolyzed Polyacrylamide in Sandstone and Carbonate Reservoirs

Abstract

Polymer flooding is an important enhanced oil recovery (EOR) method with high performance which is acceptable and applicable on a field scale but should first be evaluated through lab-scale experiments or simulation tools. Artificial intelligence techniques are strong simulation tools which can be used to evaluate the performance of polymer flooding operation. In this study, the main parameters of polymer flooding were selected as input parameters of models and collected from the literature, including: polymer concentration, salt concentration, rock type, initial oil saturation, porosity, permeability, pore volume flooding, temperature, API gravity, molecular weight of the polymer, and salinity. After that, multilayer perceptron (MLP), radial basis function, and fuzzy neural networks such as the adaptive neuro-fuzzy inference system were adopted to estimate the output EOR performance. The MLP neural network had a very high ability for prediction, with statistical parameters of R2 = 0.9990 and RMSE = 0.0002. Therefore, the proposed model can significantly help engineers to select the proper EOR methods and API gravity, salinity, permeability, porosity, and salt concentration have the greatest impact on the polymer flooding performance.

New Opportunity: Machine Learning for Polymer Materials Design and Discovery

“…Owing to the advantages of high parallelism, nonlinear global effect, good fault tolerance, self‐adaptability, and self‐learning function, ANN has been frequently applied in polymer design and discovery, industrial parameters optimization, research of phas e transition, and microstructure simulation. [ 50–54 ]…”

“…Owing to the advantages of high parallelism, nonlinear global effect, good fault tolerance, self-adaptability, and self-learning function, ANN has been frequently applied in polymer design and discovery, industrial parameters optimization, research of phase transition, and microstructure simulation. [50][51][52][53][54] Gaussian process regression (GPR) refers to a non-parametric model that uses Gaussian process to perform regression analysis of data. One of the most typical characteristics of GPR reflects on its precisely theoretical basis, which adds Gaussian prior the distribution to the non-parametric regression function and deduces the posterior distribution of the unpredicted target.…”

Recent Advances in Functional Polymer Materials for Energy, Water, and Biomedical Applications: A Review

“…A hydrogel polymer is injected into the reservoir to increase the viscosity of the fluid that contains water, making that fluid more difficult to flow than the oil, thereby increasing the production of oil. The most common polymer that is used for this application is one or more of the polyacrylamide group [113,114]. A typical polymer flood project involves the mixing and injecting of polymer over an extended period of time until about 30 to 50% of the pore volume of the reservoir has been injected.…”

“…A variety of these products is available from several manufacturers. In general, the performance of a polyacrylamide depends on its molecular weight and its degree of hydrolysis [113,114,118]. Partially hydrolyzed polyacrylamide (HPAM) is one of the polyacrylamide group, and it has the shape of a straight chain polymer of acrylamide monomers, some of which have been hydrolyzed.…”

Adaptive machine learning based surrogate modeling to accelerate PDE-constrained optimization in enhanced oil recovery

“…Artificial neural networks also gained attention in the context of enhanced oil recovery in recent years, see [40,10,2], for instance. However, these approaches mainly focus on accelerating the evaluation of the costly objective function without providing a way to solve polymer EOR optimization problems using the proposed surrogate models.…”