The traditional approach for risk acceptance criteria defined by the presence of explosive atmospheres (ATEX), consists in assessing the coincidence likelihood in time and space of both the atmosphere and the potential ignition sources.
For this criterion to work, two fundamental underlying assumptions must be considered. In first place, both the presence of the explosive atmosphere and that for the ignition source must be fully independent in its occurrence. Secondly, it must be possible to prevent the presence of ignition sources in a hazardous location during normal operational conditions.
Both assumptions are easy to apply while considering those ignition sources caused by equipment: engineering design, material specification, sensors and interlocks, etc. are some measures which can be implemented to prevent that a machine generates ignition sources while in operation, but, after all the experience gathered over the years in relation with explosive hazards, it is nowadays clear, that focussing only in equipment generated ignition sources is not enough to ensure safe processes.
It is easy to identify several types of ignition sources which are related either to process conditions or to the properties of the substances handled, not requiring any failure in Ex rated equipment. Moreover, it is also easy to show how, many times, these ignition sources are present under normal process conditions.
This paper will describe with practical examples some cases in which this conflict is clearly present and, afterwards, propose a reviewed criterion for risk acceptance of hazards defined by the presence of explosive atmospheres.