In a large pilot plant the upper explosion limit of ethene -air - nitrogen mixtures was experimented in 3.0-m-long and 21-, 50-, and 100-mm-dia. tubes at different flow rates, pressures, and temperatures. The upper explosion limit, influenced by the gas velocity, becomes smaller and shifts to higher oxygen concentrations for increasing flow rates. The results of these tubes could be correlated based on the tribe Reynolds number. A cooling effect of the tube wall, which might influence the explosion region was not observed An increase in pressure lowers the critical oxygen content as does an increase in temperature, thus the explosion region becomes larger: Different obstacles were tested, which alter the hydrodynamics. Reaction fronts could only propagate for increased oxygen concentrations through a structured Sulzer laboratory gauze packing or a sudden reduction in diameter from 50 to 20 mm. In the experiment, where the tube was completely filled with glass spheres, propagation of reaction fronts through this packed bed was not possible, even at very low gas velocities and very high oxygen concentrations. In a deep dead zone connected to the tube, the gas was ignited and reacted without igniting the gas outside the dead zone. After some time the reaction stopped because of oxygen starvation. However, dead zones with another geometry, where renewal of the combustible gas can occur, still may be dangerous.