Finding a reaction rate is one of the most essential procedure in chemical reaction engineering. This enables us to estimate not only performance of the reaction, but design of the reactor, and eventually to scale-up the plant.
Determining of reaction rate is not a simple work even
with all kinds of reaction data in various reaction conditions in hands, and
even though it follows a very good elementary reaction. The reason of this
difficulty is simply because experimental data is actually a combination of
many of physical and chemical phenomena, and we are only looking for a way to
find a reaction rate out of the data of very complex phenomena involved.
Mathematical calculation with formula, numerical calculation,
or graphical method of calculation often does not satisfy what we really want
in terms of plotting data with relations of one of the reaction variables. We
just cannot separate one phenomenon to another by looking at data itself.
Classical approach of determining a reaction rate always
requires certain assumptions; isothermal, isobaric, uniform concentration (or
perfect mixing), no residence time distribution, type of reactors and many
others. How can we possibly match these assumptions with actual data?
Plotting between variables and reaction results is a
great method to give an insight of reaction behavior. This plotting procedure I
would say is the most important work in chemical reaction engineering. At the
same time, this plot needs to have a certain tendency for engineers to
understand or utilize the meaning of Lab experiments.
To give a meaning of a trend in experimental data is a
very hard work, and not to mention determining reaction rate. Well, we can
simply forget all the conventional or classical approach of reaction
engineering for now, and let the computer work to find a relation with numbers
instead of formula.