Development and evaluation of experimental calorimetric systems for the simulation of an external heat input on reactive chemicals and monomers
Conference Paper
Overview
Identity
Additional Document Info
View All
Overview
abstract
In the chemical industry, several incidents involving the exposure of storage, transport and reactor vessels to fire have been reported. The main aim of this paper is to identify methods to improve safety measures for vessels containing reactive chemicals (e.g. monomers) that could be exposed to fire. This is done by developing and assessing experimental and theoretical methods for the measurement of the temperature and pressure rise rates resulting from a runaway reaction with external heat input. A commercially available adiabatic calorimeter was adapted for simulating the effect of an external heat input on reactive chemicals. Four heat input designs were tested. A new method of using an immersion cartridge heater with a custom test cell was shown to be the best heat input setup, the input power from the power supply being fully used to heat the system. Good experimental results were obtained with the methanol p acetic anhydride reaction (vapour system) and the decomposition reaction of 20% di-tert-butyl-peroxide in toluene (tempered hybrid system). It was measured experimentally that increasing the external heat input leads to a decrease of the reaction completion time, an increase of the maximum temperature and pressure, and an increase of the maximum temperature and pressure rise rates. This would have severe implications if it occurred in an industrial accident. The validity of two theoretical correction methods of adiabatic data were tested experimentally using the data obtained with the methanol p acetic anhydride reaction. The correction method proposed by Huff gave conservative results, with the significant advantage of only requiring limited input data. However, in the case of systems showing multiple overlapping reactions with different activation energies, Huff's approach would fail. A dynamic model taking into account the effect of external heating is likely to give better results, however its implementation would require a detailed knowledge of the kinetics of the chemical system, which is not often available. Especially when the chemical system is too complex to be simulated by a dynamic model, the kinetic data is not available, or when it is outside the application range of Huff's method, the experimental technique developed in this work would be a reliable, cost-effective and convenient alternative. 2009 Crown Copyright.
