A5. Control and consequences of small foreseeable release scenarios

The small foreseeable release experiments were aimed to examine the effect of the vent area in function of the pressure during deflagrations of the explosive mixtures, (hydrogen/air or methane/air), in uniform and non-uniform conditions. Vented explosion experiments have been carried out in a cubic structure made of steel section bars with an internal volume of 25 m3.

The purpose of the experimental work performed in sub-task 5.5 by the University of Pisa (UNIPI) was to fix the size of Zone 2 ATEX in case a small accidental leakage of the fuel cell occurs. In this case, the limit of H2 in air is 2% H2 vol. There were five sensors of H2 concentration to define the volume around the fuel cell where the limit is not overtaken.
The most credible loss of H2 is located on the valve of the inlet gas pipeline, before the pressure reducer. The internal pressure between 2 bar and 5 bar has been considered, after the pressure reducer (inside PentaH2) the pressure value decreases to 350 mbar.  In order to take into account the worse case of loss, the highest 5 bar pressure value has been chosen. The calculations of the H2 flow GH2 has been performed with EFFECT-SGIS 7.3.

The diameter of the leakage has been chosen according to ATEX limits in order to calculate Zone 2.In the case of pipelines with diameters of up to 150 mm, the guide refers to a small accidental leakage from a valve (exactly the case of the H2 inlet pipeline). The ATEX loss value from a valve is analogous to a flux from a hole with the diameter ΦL = 0.56 mm i.e. area AL = 0.25 mm2.
Calculations are referred also to areas of 0.5 mm2 and 1 mm2 to get the opportunity to study more dangerous cases then the ATEX one.

The correct ventilation in an enclosure in order not to exceed the concentration of 2% depends on several factors. These factors are the geometry of the vents, their size, the strength of the wind, the gradient of temperature between the internal part of the enclosure and the external one.
In natural ventilation (NV) experiments, the worst conditions of ventilation were considered concerning the wind and the temperature. In fact most of the tests have been carried out with wind ≈ 3 m/s or less and internal temperature ≈ 25°c. Where specified, a small number of tests have been performed with wind 5 – 6 m/s and gradient of temperature of 10°c (internal 40°c, external 30°c).

The NV main result is that it is easy to prevent the accumulation of H2 in the enclosure in various geometries considering the ATEX leak (no more than 40 l/min).

The forced ventilation (FV) tests were conceived to complete the evidence belonging to the natural ventilation experiments. FV is carried out on the CVE facility using two industrial fans. Using one fan in aspiration, it is enough to ensure the evacuation of the hydrogen, maintaining the concentration below 2% considering leakage flow till 90 l/min.

 The analysis of the natural and forced ventilation efficiency leads to the conclusion that safety systems should be adopted for all enclosures where a credible non-catastrophic leakage can occur: To calculate the vent area, it is convenient to choose the geometry first, then the leak size and consequently the size of vent areas that are shown in Table 1. The result is the minimum vent area in that geometry for which the enclosure meets the ATEX limit of zone 2 (H2 concentration no more than 2% vol.). Where possible, it is convenient to use one or more suitable solutions, for example:

• To reasonably increase the vent areas beyond the minimum value in Table 1;
• To consider the vent areas for a leak flow reasonably bigger than the minimum;
• To incline the roof, making the NV easy and efficient;
• To install a small fan able to remove the internal mixture from the enclosure.
• The limit of 40 l/min for the leak is suitable for all types of fuel cell in civil use, it is reasonable to consider leaks no larger than 90 l/min because this is the largest reasonable value for a catastrophic leakage.