Complex oxide (slag) systems play a key role in a wide range of important industrial
applications, including metallurgical and coal utilisation processes. To be able to predict the
behaviour of the mineral matter at high temperatures during coal combustion/gasification or
metallurgical processes it is essential to know the presence, proportion, compositions and
physico-chemical properties of the liquid slag and the solid phases as a function of bulk
composition and operational parameters, such as temperature and atmosphere
(reducing/oxidising). Even small improvements to operating practice can result in substantial
economical advantages due to significant scales of the industries.
The aims of the present study are:
• to analyse and improve (if possible) the existing thermodynamic solution models;
• to develop a software for phase equilibrium calculations, which can be further used for
incorporating modified thermodynamic models in phase equilibrium calculations and
for facilitating the development of thermodynamic databases;
• to demonstrate the applicability of the approach, which involves thermodynamic and
viscosity modelling, for predicting the behaviour of the mineral matter in industrial
• to develop customer-oriented software that can assist operational engineers and system
designers in the prediction of the mineral matter behaviour.
The following work is done to achieve these goals.
Basic thermodynamic solution models, the process of thermodynamic optimisation and
leading software and database packages are reviewed (Chapter 1).
Features of the quasichemical model that are specific for a binary system with strong short
range ordering are analysed using the re-optimisation of the system K2O-SiO2 as a working
example (Chapter 2). The liquidus temperature calculator is developed. The developed
calculator is used as a tool for selecting adjustable parameters of the quasichemical model.
The quasichemical solution model is further critically analysed (Chapter 3). Shortcomings
of the recent modifications of the model are demonstrated. An interesting effect related to
asymmetric methods of extrapolation of binary interaction parameters into multicomponent
systems is also demonstrated.
Geometrical methods of extrapolation of binary interaction parameters into ternary and
multicomponent systems are analysed (Chapter 4). A probabilistic interpretation of the
parameters of the polynomial solution model is suggested. The interpretation provides a
theoretical justification of the power series multicomponent polynomial models and a way inwhich these models can be incorporated into the Calphad approach.
A software for phase equilibrium calculations is developed (Chapter 5). The developed
software enable incorporating modified thermodynamic models in phase equilibrium
predictions. The ability to calculate the viscosity of liquid slags in accordance with the
quasichemical viscosity model is also incorporated in the developed software.
Thermodynamic and viscosity modelling are applied for predicting Ash Fusion
Temperatures (AFT), which are used in industrial and commercial practice to assess coal ash
fusibility (Chapter6). The pooled variance statistical technique is used for estimating the
internal variability of AFT datasets, which is important for the quality assessment of different
sets of predictors. Software for predicting AFT of coal blends is developed.
The applicability of the phase equilibrium and viscosity modelling for the prediction of the
adhesion tendency of heterogeneous coal ash particles is demonstrated (Chapter 7). Two factors
influencing the sticking probability, namely the proportion of liquid phase in a heterogeneous
particle and the viscosity of the liquid, are analysed.