Mixture Reactivity Effects on Explosion Venting

Free vented explosions were investigated for 10% methane, 4.2 and 4.5% propane, 6.5 and 7.5% ethylene, 30% and 40% hydrogen in a 10 litre cylindrical explosion vessel for vent coefficients of 4.3 and 21.7. The cylindrical vessel volume was 10L and had a diameter of 162mm and an L/D of 2.8. End ignition was used on the wall opposite the vent. The results are presented against KG and the laminar burning velocity as measures of the mixture reactivity. It is shown that the correlation of the KG effect by Bartknecht does not agree with other experimental data, although the hydrogen results are closer to the present results than the other gases. In contrast the laminar flame venting theory, as used in NFPA68 (2013), does correlate the data well, even though it is not supposed to apply to hydrogen explosions. There was evidence of very fast flames at the vent for hydrogen explosions. Acceleration of the flames towards the vent was demonstrated, due to the expansion of the burnt gases in the direction of the vent. The laminar flame venting theory that is used in NFPA68 (2013) over predicts the measured Pred due to the assumption of the vessel surface area as the area of the flame at Pred. It was shown that the flame arrives at the wall after the flame has vented the vessel and well after the time that Pred occurs. At Kv 4.3 the external overpressure was responsible for Pred, although the difference from Pfv was small for methane, propane and ethylene but for hydrogen the flow through the vent Pfv was the highest overpressure. At Kv = 21.7 the pressure loss due to the unburnt gas flow through the vent was the largest overpressure. For hydrogen sonic flow at the vent occurs and at high Kv sonic flow is predicted to occur using the laminar flame venting equation modified for sonic flow at the vent. Sonic flow at the vent is not taken into account in current venting guidance.