The Catalyst Review April 2020 - 9

SPECIAL FEATURE
economically. The response 25 years ago from I.G. Snyder, director of process technology development at Dow Chemicals, is that: "In
order for a company to remain world-class competitive, it will now be necessary to run the plant with less maintenance, higher yield,
less energy, product quality at least as good or better than before, and at the same time, increased productivity, which means fewer
people and less capital. To realize this level of performance, it is no longer enough to optimize individual unit operations, but rather it
is now necessary to simulate and optimize the entire process and plant. This requires, a new level of understanding, which includes a
knowledge of the basic reaction and transport mechanisms in the catalyst and reactor. If entire processes are understood at this level, it
could be possible to increase the output of most of our reactor systems as much as 50-to-100%" (Krieger, 1995).
Procedures in Scale-Up of a Catalytic Reaction from the Lab to the Working Plant
The procedures in scale-up of a catalytic reaction are diagrammed in Figure
1. The first task in scale-up of a catalytic process begins with experimental
design (Figure 1). The most effective approach is a sophisticated statistical
design of experiments obtained with the assistance of a statistician. Rather
than a full n-factorial experiment, a partial n-factorial design with analysis
of interactions between variables enables the reduction of the number of
experiments by a factor of 2-3. Minimizing the number of experiments is
especially important for long runs, e.g., typical deactivation experiments
that require operation over several hundred to several thousand hours. For
example, the dissertation work of Paul (2008) utilized sequential D-optimal
design for stepwise minimization of errors in reaction rate constants for
Fischer-Tropsch; in this method, the previous results enable the selection of
the optimal set of conditions for the next experiment. However, acceptable
minimization of errors in rate constants, activation energies, and adsorption
equilibrium constants requires measurements over wide ranges of reactant
concentrations and temperature to produce a minimum of 25-30 data points
plus replicates.

Figure 1. Components, methods, and scientific disciplines
involved in scale-up of a catalytic process from laboratory to the
full-scale plant.

The second step is activity/selectivity measurements carried out according to
the experimental design, using a differential reactor (run at low conversion
and near constant reaction rate). The reactor can be a fixed bed containing
0.1 g of catalyst or a CSTR with a few grams of catalyst but in either case
with state-of-the-art controls of reactor temperature and reactant and
product flowrates. The resulting intrinsic rate data, obtained in the absence
of heat and mass transport limitations are converted to intrinsic activity and
selectivity data, which in turn, are inputs to macro-kinetic rate expressions
Source: Adapted from Smith 1968.
and selectivity models. From theoretical DFT calculations and transition
state theory, rate constants and activation energies can be calculated for
each postulated elementary step using intrinsic experimental activity data as constraints. Comparison of the constants from macro- and
micro-kinetic models can be used to optimize rate constants, activation energies and heats of adsorption in the intrinsic macro-kinetic
model or rate expression for each stoichiometric reaction. Where rate or sintering data are fitted to a macro-kinetic model, use of a
nonlinear correlation coefficient to establish goodness of fit is the best practice.
The third step is the development and validation of a reactor model which includes: (1) mass, energy, and momentum balances;
(2) catalyst physical and geometric properties; (3) rates of heat and mass transfer across catalyst particle surface interfaces; and (4)
diffusion of reactants and products through catalyst pores. A comprehensive, representative model relies on high quality representative,
intrinsic rate data and sufficiently sophisticated models of heat and mass transport, and pore diffusion. Examples of comprehensive
models are discussed in the next section.
The reactor model is combined with reactor and plant design fundamentals in Step 4 to design the pilot catalytic reactor and plant.
Implicit in this step is also construction of the pilot plant and experimental design for operating the pilot plant. Data from the operation
of the pilot plant will be used to validate and refine the reactor model.
Pilot plant data and the refined reactor model and experience running the pilot plant are combined with process design and economic
fundamentals in Step 5 to design and construct the full-scale plant. The objectives are to minimize the capital and operating costs of the
plant and maximize profit of the products produced by the plant. However, Objective 2 is the principal objective function for optimizing
the design of the plant. Trade-offs may be worthwhile, e.g., investing in a more sophisticated design with a higher capital cost in order to
maximize the overall productivity and profit in the long run.
The Catalyst Review

April 2020

The Catalyst Review April 2020

Table of Contents for the Digital Edition of The Catalyst Review April 2020