21st European Symposium on Computer Aided Process Engineering – ESCAPE 21
E.N. Pistikopoulos, M.C. Georgiadis and A.C. Kokossis (Editors)
© 2011 Elsevier B.V. All rights reserved.
Stochastic Monte Carlo Simulations as an Efficient
Multi-Scale Modeling Tool for the Prediction of
Multi-Variate Distributions
Dimitrios Meimaroglou,
a,c
Costas Kiparissides
a,b
a
Chemical Process Engineering Research Institute, Centre for Research and
Technology Hellas, Thessaloniki, Greece 570 01
b
Department of Chemical Engineering, Aristotle University of Thessaloniki,
Thessaloniki, Greece, 60361
c
Current Address: Laboratory of Reactions and Process Engineering, Nancy University,
LRGP-ENSIC-INPL, 1 rue Grandville, BP 20451, 54000 Nancy, France
Abstract
In the present work, the role of stochastic Monte Carlo simulations as an efficient
computational tool for the simulation of multi-variate physico-chemical systems, within
the general framework of population balances, is presented. The applicability of the
Monte Carlo method to different length scales of a process is clearly illustrated in two
representative examples, namely the prediction of the bi-variate particle size distribution
of particulate processes and the calculation of distributed molecular properties of
highly-branched polymers on the basis of their exact topological architecture.
Keywords: stochastic simulations, Monte Carlo, population balance equation, particle
size distribution, highly-branched polymers.
1. Introduction
Mathematical modeling and in silico analysis of chemical processes have been widely
used for more than 40 years and have contributed to great advances in numerous
research areas. Depending on the complexity of the process under study, the
mathematical model implemented for its description may contain systems of algebraic,
differential or integro-differential equations of different order. Accordingly, a number of
advanced numerical methods have been developed to deal with the complex systems of
equations a mathematical model may contain. These methods are broadly classified into
two general categories, namely the deterministic and stochastic numerical methods.
The deterministic approach is based on the principle that events are bound by causality,
i.e., any state of the system under study is majorly determined by prior states. This
approach is commonly characterized by low computational requirements (for simple
systems) and high accuracy and repeatability in the calculated results. On the other
hand, its application to multi-dimensional process models is bounded by an increased
level of complexity.
An alternative approach to the commonly employed deterministic numerical methods is
the use of probabilistic tools (i.e., Monte Carlo simulations). The stochastic approach is
based on the principle that a system's subsequent state is mainly determined by a
random element. Stochastic methods have attracted significant attention over the last