Fire and Explosion in an Alcohol Tank: An Opportunity for a Systemic Vision for Process safety in a Sugar and Alcohol Industry 

Understanding the risk of explosions for flammable and combustible fuels is a key factor when performing risk assessments and siting studies for facilities. However, there are a number of fluids (e.g., ammonia, new climate-preserving refrigerants) that have been characterized as mildly flammable, whereby the laminar burning velocity (LBV) is less than 10 cm/s. In comparison, more traditional extremely flammable gases (e.g., propane and methane) have LBVs of approximately 40 cm/s. A very conservative approach to assess the risk with such fuels, is to assume the explosion properties of such mildly flammable fuels are close to those of methane when evaluating the potential explosion consequences. This will, however, grossly over-predict the potential explosion consequences as the resulting flame speed and overpressures are directly correlated to the LBV of the fuel. At the opposite end of the spectrum, another approach is to essentially assume that there is no risk from such fuels.

In order to properly evaluate the explosion risk of such mildly flammable fuels on the large scale, the flammability characteristics must be thoroughly understood so that the probability and consequence of fire/explosion hazards can be analyzed and compared to the risk of extremely flammable fluids. However, a thorough understanding of the flammability and explosivity characteristics for these mildly flammable fluids has not been performed at either lab and test scales. The results of such a study would enable facilities to better design against such risks and help to optimize appropriate mitigation measures, especially if consequence tools such as CFD are updated and validated against such data.

Large-scale experiments were performed to evaluate deflagration consequences of such compounds like ammonia at a more practical scale (50 m³). These tests were performed in a partially confined environment with two degrees of obstacle density (congestion) placed within the flame path. These tests were also performed with methane to provide a direct comparison of the resulting deflagration consequences. Supplementing the large scale test results, the present study also measured the following properties for such fuels like ammonia: upper and lower flammability limits; auto-ignition temperature; minimum ignition energy; explosion severity index; peak pressure rise; and burning velocity as a function of equivalence ratio. The results are compared against the limited data available in literature and to those of methane. In addition, CFD validation tests were performed with the new large-scale experimental data, and very good agreement was obtained for the low LBV fuels. The combined large-scale consequence testing, fundamental flammability and ignitibility experiments, as well as modeling results will allow for direct comparisons of risk.