Quantifying the uncertainty introduced by discretization and timeaveraging in twofluid model predictions
The twofluid model (TFM) has become a tool for the design and troubleshooting of industrial fluidized bed reactors. To use TFM for scale up with confidence, the uncertainty in its predictions must be quantified. Here, we study two sources of uncertainty: discretization and timeaveraging. First, we show that successive grid refinement may not yield gridindependent transient quantities, including crosssection–averaged quantities. Successive grid refinement would yield gridindependent timeaveraged quantities on sufficiently fine grids. A Richardson extrapolation can then be used to estimate the discretization error, and the grid convergence index gives an estimate of the uncertainty. Richardson extrapolation may not work for industrialscale simulations that use coarse grids. We present an alternative method for coarse grids and assess its ability to estimate the discretization error. Second, we assess two methods (autocorrelation and binning) and find that the autocorrelation method is more reliable for estimating the uncertainty introduced by timeaveraging TFM data.
 Authors:

^{[1]}
;
^{[2]};
^{[3]}
 National Energy Technology Lab. (NETL), Morgantown, WV (United States). Science & Technology Strategic Plans & Programs
 National Energy Technology Lab. (NETL), Morgantown, WV (United States). Research & Innovation Center; West Virginia Univ., Morgantown, WV (United States). Mechanical and Aerospace Engineering
 National Energy Technology Lab. (NETL), Morgantown, WV (United States). Research & Innovation Center
 Publication Date:
 Report Number(s):
 NETLPUB21134
Journal ID: ISSN 00011541
 Type:
 Accepted Manuscript
 Journal Name:
 AIChE Journal
 Additional Journal Information:
 Journal Volume: 63; Journal Issue: 12; Journal ID: ISSN 00011541
 Publisher:
 American Institute of Chemical Engineers
 Research Org:
 National Energy Technology Lab. (NETL), Pittsburgh, PA, and Morgantown, WV (United States)
 Sponsoring Org:
 USDOE Office of Fossil Energy (FE)
 Country of Publication:
 United States
 Language:
 English
 Subject:
 42 ENGINEERING; 22 GENERAL STUDIES OF NUCLEAR REACTORS; two‐fluid model; multiphase computational fluid dynamics; uncertainty quantification; discretization error; time‐averaging error
 OSTI Identifier:
 1440343
Syamlal, Madhava, Celik, Ismail B., and Benyahia, Sofiane. Quantifying the uncertainty introduced by discretization and timeaveraging in twofluid model predictions. United States: N. p.,
Web. doi:10.1002/aic.15868.
Syamlal, Madhava, Celik, Ismail B., & Benyahia, Sofiane. Quantifying the uncertainty introduced by discretization and timeaveraging in twofluid model predictions. United States. doi:10.1002/aic.15868.
Syamlal, Madhava, Celik, Ismail B., and Benyahia, Sofiane. 2017.
"Quantifying the uncertainty introduced by discretization and timeaveraging in twofluid model predictions". United States.
doi:10.1002/aic.15868. https://www.osti.gov/servlets/purl/1440343.
@article{osti_1440343,
title = {Quantifying the uncertainty introduced by discretization and timeaveraging in twofluid model predictions},
author = {Syamlal, Madhava and Celik, Ismail B. and Benyahia, Sofiane},
abstractNote = {The twofluid model (TFM) has become a tool for the design and troubleshooting of industrial fluidized bed reactors. To use TFM for scale up with confidence, the uncertainty in its predictions must be quantified. Here, we study two sources of uncertainty: discretization and timeaveraging. First, we show that successive grid refinement may not yield gridindependent transient quantities, including crosssection–averaged quantities. Successive grid refinement would yield gridindependent timeaveraged quantities on sufficiently fine grids. A Richardson extrapolation can then be used to estimate the discretization error, and the grid convergence index gives an estimate of the uncertainty. Richardson extrapolation may not work for industrialscale simulations that use coarse grids. We present an alternative method for coarse grids and assess its ability to estimate the discretization error. Second, we assess two methods (autocorrelation and binning) and find that the autocorrelation method is more reliable for estimating the uncertainty introduced by timeaveraging TFM data.},
doi = {10.1002/aic.15868},
journal = {AIChE Journal},
number = 12,
volume = 63,
place = {United States},
year = {2017},
month = {7}
}