(65a) A Simplified Approach to Gas Explosion Vent Sizing

Accidental explosions within buildings and pieces of equipment can cause significant damage to the enclosure itself as well as the surrounding area. To mitigate the damage caused by these events, explosion venting is commonly employed. Proper implementation of explosion venting, however, requires that the vent area is appropriately sized to the hazard and enclosure.

Many parameters, such as the enclosure volume, location of the vent, reactivity of the fuel, presence of obstacles, and the performance of the vent panel itself can significantly affect the vent area required to reduce pressures below the design strength of an enclosure. Existing correlations, including the current NFPA 68 and EN 14994 standards have been found to both significantly under and over predict the vent area required in many cases. To address the deficiencies of existing correlations, a new, physics-based, multi-peak model was developed in previous work. This detailed model predicts the peak pressures at the actual test conditions, allowing for an accurate comparison across the full set of experimental data available in the literature, not just experiments performed in a worst-case configuration.

Deployment of detailed models for actual engineering applications, however, present a number of challenges. To account for all of the parameters described above, detailed models can be extremely complex, requiring a large number of inputs and solving several coupled equations, often iteratively. Due to the complexity of these models, they are difficult to accurately implement and are often sensitive to parameters that are challenging or even impossible to quantify for real geometries. For a consistent design and application of explosion venting, there is a great need for simplified models with reduced complexity and a smaller set of input parameters.

In this study, the previously developed detailed model is exercised to determine the key parameters that affect the overall peak pressure generated by vented explosions. The model is then simplified, reducing the number and complexity of the formulation, by isolating the key parameters that affect pressure generation in realistic geometries. Next, the inputs to the model are generalized into a reduced set of inputs that can be easily obtained without requiring detailed measurements of individual components within an enclosure.

This simplified model is then compared with an extensive set of experimental test data from the literature, with careful attention to consider which experiments, and individual pressure peaks, are not consistent with real world configurations. The overall performance of the simplified model is then assessed through comparison with existing vent sizing correlations.