From Experimental Data Via Kinetic Model to Predicting Reactivity and Assessing Reaction Hazards

Abstract

There is no other way to get original data regarding chemical reaction but experimental study. What kind of experimental technique to use depends on the aim of a study. Nevertheless, there is one almost universal method that is particularly applicable in such areas as predicting reactivity, assessment of reaction hazards, thermal stability of chemicals, etc. This is calorimetry of various types (DSC, isothermal, reaction, adiabatic etc.). Very often calorimetry is used for direct experimental determination of necessary characteristics. At the same time, introduction of mathematical simulation can provide obtaining much more versatile and reliable results in combination with more complete use of information contained in experimental data. The approach to investigating chemical reaction and predicting its behavior under various conditions, which systematically uses kinetics-based simulation, is the subject of the paper. At first, several examples are shown that demonstrate efficiency of the approach. The following cases are discussed: • Predicting adiabatic course of a reaction using DSC and adiabatic data • Analyzing thermal stability of a product, influence of product's composition and presence of contaminants • Determining reactivity rating number of a chemical product • Determining critical parameters of thermal explosion • Simulating runaway in a BATCH and vent sizing • Designing an inherently safer process All these examples are based on real experimental data and regard some typical practical problems. Then the general scheme of the simulation-based approach is discussed which consists of three basic steps: • Carrying out the necessary set of calorimetric experiments and proper processing of data for kinetics evaluation; • Creating the mathematical model of a reaction- evaluating kinetics; • Solving practical problem using mathematical (numerical) simulation. The main merits of the approach are as follows • possibility to apply more adequate complex mathematical models of processes; • possibility to simulate and analyze various scenarios of process proceeding; • possibility to model thermal explosions and runaways without essential simplifications; • principal solution of the scale-up problem. Some problems dealing with practical application of the proposed method are examined. In conclusion, the composition of a problem-oriented software series is discussed that simplifies introduction of simulation-based methods into research practice. The programs of this series form three groups intended for initial processing of calorimetric data, creation of mathematical model of a reaction, and for simulation of processes' proceeding.