A stochastic well-test analysis on transient pressure data using iterative ensemble Kalman filter

Bazargan, Hamid; Adibifard, Meisam

doi:10.1007/s00521-017-3264-5

Cited by 10 publications

(5 citation statements)

References 64 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Ensemble data assimilation (EnDA) [43,44] methods have shown a strong performance in dealing with non-linear chaotic DA systems by creating an ensemble with size M of the system state depicted as {x (i)…”

Section: Ensemble Methodsmentioning

confidence: 99%

Observation Error Covariance Specification in Dynamical Systems for Data assimilation using Recurrent Neural Networks

Cheng¹,

Qiu²

2021

Preprint

View full text Add to dashboard Cite

Data assimilation techniques are widely used to predict complex dynamical systems with uncertainties, based on time-series observation data. Error covariance matrices modelling is an important element in data assimilation algorithms which can considerably impact the forecasting accuracy. The estimation of these covariances, which usually relies on empirical assumptions and physical constraints, is often imprecise and computationally expensive especially for systems of large dimension. In this work, we propose a data-driven approach based on long short term memory (LSTM) recurrent neural networks (RNN) to improve both the accuracy and the efficiency of observation covariance specification in data assimilation for dynamical systems. Learning the covariance matrix from observed/simulated time-series data, the proposed approach does not require any knowledge or assumption about prior error distribution, unlike classical posterior tuning methods. We have compared the novel approach with two state-of-the-art covariance tuning algorithms, namely DI01 and D05, first in a Lorenz dynamical system and then in a 2D shallow water twin experiments framework with different covariance parameterization using ensemble assimilation. This novel method shows significant advantages in observation covariance specification, assimilation accuracy and computational efficiency.

show abstract

Section: Ensemble Methodsmentioning

confidence: 99%

Observation Error Covariance Specification in Dynamical Systems for Data assimilation using Recurrent Neural Networks

Cheng¹,

Qiu²

2021

Preprint

View full text Add to dashboard Cite

show abstract

“…In this case, evolutionary and stochastic algorithms could be a promising solution for different applications. For the application of LSW ood in sandstone reservoirs, many studies have been conducted to determine optimization parameters [12,[21][22][23]. Based on the conducted studies, optimization studies to determine relative permeability and capillary pressure functions using the combination of evolutionary and machine learning algorithms are rare.…”

Section: Full Textmentioning

confidence: 99%

Hybrid Particle Swarm Optimization Coupled by Machine Learning Technique for Estimating Relative Permeability and Capillary Pressure Functions During Low-Salinity Flood

Khosravi,

Simjoo,

Chahardowli

2024

Preprint

View full text Add to dashboard Cite

Experimental methods to estimate the relative permeability and capillary pressure data are expensive and time-consuming, thus evolutionary and stochastic algorithms to estimate these parameters could be a promising solution. This study aims to determine the relative permeability and capillary pressure functions of a heterogeneous sandstone core in the presence and absence of clay during low-salinity water flood. The relative permeability and capillary pressure data were determined by automatic history matching recovery factor and pressure drop results from previously lab-reported data, through coupling a simulator with the particle swarm optimization algorithm. A series of correlations were proposed based on multiple-linear regression for relative permeability parameters at the desired low-salinity conditions. They were validated against experimental results of no clay and clayey formation with regression of 95% and 97%, and MSE of 1.2E-04 and 8E-05. One single curve for relative permeability and capillary pressure obtained from averaging methods was assigned to the grid cells for core scale low-salinity flood simulation. Accordingly, the effect of salinity and clay content on the obtained curves was investigated. The proposed hybrid method could be a suitable tool to estimate relative permeability and capillary pressure functions at the core and field level of the water-based EOR methods.

show abstract

“…In this case, evolutionary and stochastic algorithms could be a promising solution for different applications. For the application of LSW flood in sandstone reservoirs, many studies have been conducted to determine optimization parameters 12 , 21 – 23 . Among them, optimization studies to determine relative permeability and capillary pressure functions using the combination of evolutionary and machine learning algorithms are rare.…”

Section: Introductionmentioning

confidence: 99%

Low salinity water flooding: estimating relative permeability and capillary pressure using coupling of particle swarm optimization and machine learning technique

Khosravi,

Simjoo,

Chahardowli

2024

Sci Rep

View full text Add to dashboard Cite

The reservoir’s properties are required for proper reservoir simulation, which also involves uncertainties. Experimental methods to estimate the relative permeability and capillary pressure data are expensive and time-consuming. This study aims to determine the relative permeability and capillary pressure functions of a sandstone core in the presence and absence of clay during low-salinity water floods. The data were provided by automatic history matching the results from previously lab-reported studies through coupling a simulator with the particle swarm optimization algorithm. Correlations were proposed using multiple-linear regression for relative permeability and capillary pressure parameters at low-salinity conditions. They were validated against experimental results of no clay and clayey formation with regression of 95% and 97%. To assign one curve of relative permeability and capillary pressure to the grid cells of the simulator, averaging techniques were implemented. The effect of salinity and clay content on the obtained curves was investigated. Changing salinity from 42000 to 4000 ppm, the reduction in water relative permeability appeared to be higher than the oil relative permeability increment. Moreover, a noticeable shift in the relative permeability curves toward the highest saturations related to the clay content was observed. The proposed hybrid method could be a suitable tool to estimate the relative permeability and capillary pressure functions of the water-based EOR methods.

show abstract

A stochastic well-test analysis on transient pressure data using iterative ensemble Kalman filter

Cited by 10 publications

References 64 publications

Observation Error Covariance Specification in Dynamical Systems for Data assimilation using Recurrent Neural Networks

Observation Error Covariance Specification in Dynamical Systems for Data assimilation using Recurrent Neural Networks

Hybrid Particle Swarm Optimization Coupled by Machine Learning Technique for Estimating Relative Permeability and Capillary Pressure Functions During Low-Salinity Flood

Low salinity water flooding: estimating relative permeability and capillary pressure using coupling of particle swarm optimization and machine learning technique

Contact Info

Product

Resources

About