Uncertainty Quantification (UQ) is the statistical analysis of how model inputs and assumptions affect the results of a simulation. This is most commonly presented as confidence intervals around results, but it also applies to parameter calibration and assessing model bias. UQ is particularly important when model verification is difficult because experimental results are unavailable, there is uncertainty in input physical parameters, or results are transient in nature. The CEL has spent significant effort applying UQ to sub-scale models of solid sorbent carbon capture systems as part of the Lab’s work with the DOE's Carbon Capture Simulation Initiative (CCSI). This includes collaborating with Los Alamos Laboratories to implement the Bayesian Spline Smoothing Analysis of Variance (BSS-ANOVA) framework on hydrodynamic and thermodynamic simulations of multiphase bubbling fluidized beds. The ultimate goal of this research is to develop full-scale models based on filtered fine-grid sub-scale simulations, but this can only be reliably achieved once model uncertainties are well understood.

• C.B. Storlie, W.A. Lane, E.M. Ryan. (accepted). Calibration of Computational Models with Categorical Parameters and Correlated Outputs via Bayesian Smoothing Spline ANOVA. Journal of the American Statistical Association.
• W.A. Lane, C.B. Storlie, C.J. Montgomery, E.M. Ryan. (2014). Numerical modeling and Bayesian calibration of a bubbling fluidized bed with immersed horizontal tubes. Powder Technology, 253: 733-743.