D6B.1A & D6B.1B Inventory, modeling and experiments related to ventilation in different types of pressure reducing stations in the distribution (low pressure) grid in the Netherlands, in natural gas and hydrogen atmospheres

This research report is a follow-up to the gas stations work package from Hydelta 1.0. That study showed that for various types of gas cabinets, hydrogen more often leads to a combustible mixture at the ventilation openings than natural gas, assuming the leakage flow rates chosen in that study. This led to the recommendation to carry out additional research to identify the effects of smaller, more common leakage flow rates. It was also recommended to investigate which types of different gas cabinets are frequently used in the Netherlands. This report further develops these recommendations.

This follow-up research is important because ventilation is an important measure in the event of an unintentional gas leak. Ventilation dilutes the gas and minimizes the risk of ignition or explosion. For the transition to hydrogen, it is important to know whether these gas pressure regulating stations with the same types of gas cabinets carry the same risk with the application of hydrogen.

The aim of this study is to gain further insight into how hydrogen behaves in existing gas cabinets compared to natural gas. That insight was obtained by looking at the issue from different angles: experiments as well as simulations using finite element method (CFD). This provided interesting insights that will help policymakers determine whether, and if so what, further measures can be taken.

1st step: Inventory common types of gas cabinets

Some 55,000 gas stations are operated by district system operators (DSOs) in the Netherlands. An inventory of gas cabinets used in the Netherlands was carried out. It focused on cabinets installed by DSOs in the last 10 years, because conversion of installations to hydrogen will initially take place with relatively new cabinets and these installations are designed in accordance with NEN1059. This does not mean they are exactly the same, but they are designed with the same minimum functional requirements. Three types of gas cabinets comprise a substantial part of the total amount of stations. These are mini-gas cabinets with a volume of: less than 0.5 m³ (these are mostly used for a high-pressure delivery station), ½ m³ gas cabinets and 4 m³ gas cabinet stations.

What is a realistic leakage size?

There has been a lot of focus in this study on which leakage is representative in (normal) operation and for which leakage rate ventilation should be effective. Different sources use different assumptions to determine the expected leakage rate. This is not surprising, as leakage rates can differ due to operational pressures, maintenance or environmental factors.

This study tested both with the highest leakage rate from a recent field study (40 l/h) of more than 700 gas stations. Also, leakage flow rates were based on leakage openings in other standards (0.025 mm² and 0.25 mm²). This is still a wide range of leakages where, especially for the larger leaks, it is expected to be noticed by the public coincidentally in close proximity of the gas cabinet with a leak. The leaks measured the field study of more than 700 stations are considered realistic, with the largest measured leak being 40 liters per hour.

2nd step: Measurements

An extensive test program was carried out with a ½ m³ cabinet and a 4 m³ cabinet station. With a mini-cabinet, some indicative measurements were performed to get a first impression. This was done by positioning a reference leak in the center of the cabinet during the experiments. From this, gas (hydrogen or natural gas) flows at a known flow rate controlled by a Mass Flow Controller. The gas concentration was then measured at various points in the gas cabinet, directly at the vent openings outside the gas cabinet and at a distance of about 0.5 meter away from the gas cabinet. The smallest leak selected has a flow rate of 40 l/h of natural gas (125 l/h for hydrogen). The largest leak is based on a leak opening of 0.25 mm² at a pressure of 8 bar (that is: 1.8 m³_n /h natural gas or 5.6 m³_n /h hydrogen). Between these extremes, several other leakage flow rates were chosen.

Key data from the experiments are shown in this document (the graphs can be found in the full document) for the ½ m³ cabinet, the 4 m³ cabinet station and the mini-cabinet where the leakage rate on the x-axis decreases in size. The first dataset on the x-axis represents both 1.8 m³_n /h natural gas and 5.6 m³_n /h hydrogen. This reasoning also applies to the other, smaller leakage flow rates.

In the case of the mini-cabinet, some indicative measurements were done to check for leakages at a leakage opening of 0.025 mm² at 8 bar pressure (i.e. 0.18 m³_n /h natural gas or 0.56 m³_n /h hydrogen) and 1 bar pressure (i.e. 40 l/h natural gas or 125 l/h hydrogen). 

When the measurements are examined more closely, the following can be concluded:

Measurement results ½ m³ cabinet

• For a 40 l/h natural gas leak, the maximum gas concentration in a ½ m³ cabinet is about 1 vol%, a mixture below the lower flammability limit (the LEL/LFL of hydrogen considered in this study is 4 vol% in air). For a similar leak for hydrogen (125l/h), the concentration rises to 2.7 vol%, which is also below the lower flammability limit. Directly at the vent openings, similar concentrations are measured, with 3.2 vol% as a local, time-dependent outlier. Half a meter away from the gas cabinet, the concentration is well below the lower flammability limit in all cases.
• Measurements with leakage openings of 0.25 mm² and 0.025 mm² measured flammable mixtures with both natural gas and hydrogen.

Measurement results 4 m³ cabinet station

• For both natural gas (40l/h) and hydrogen (125 l/h), the concentration remains well below the lower flammability limit. The measured concentrations with hydrogen are higher, but in all cases below the limit of a combustible mixture.
• In measurements with leak openings of 0.25 mm² and 0.025 mm², no combustible mixture was measured with natural gas. In the case of hydrogen, a combustible mixture was measured at leak openings of 0.25 mm².

Measurement results mini-cabinet

• Some measurements were carried out at the mini-cabinet. These aimed to provide input to any follow-up research and are too limited to draw firm conclusions.
• An actual observation is that no combustible mixture was measured with natural gas (40l/h), while it was measured with hydrogen (125l/h).

3rd step: verification with CFD calculations

As verification and to make measurements visual, CFD calculations were carried out. The aim of these modelling activities is to understand the factors affecting the flow phenomena of hydrogen gas in gas stations. Measurement data from Hydelta 1.0 were used as a reference for the mathematical models. The quality of the CFD validation is not (yet) so high that the results can be used independently of the field measurements. This is therefore an additional tool. Several reasons can be given, such as the effect of wind (variable wind) or differences in temperature (in different experiments).

The main lesson to be drawn from the CFD calculations is the influence of the roof configuration. The ventilation path is important and should facilitate the upward flow of hydrogen driven by the buoyancy. For the modelled gas cabinet with the overhanging roof lid, it can be seen that mainly hydrogen is limited to escape from the gas cabinet by a "siphon" effect. The buoyancy of hydrogen is insufficient to overcome the hydraulic resistance. An alternative CFD geometry where the overhang of the gas cabinet lid is eliminated confirms a significant improvement in ventilation.

Insights gained:

If hydrogen is transported using gas pressure regulating stations and the existing gas cabinets, it is expected that in case of a leakage, the concentrations in the gas cabinet will be higher than with natural gas. This study also looked at possible modifications and their effect. These are modifications that can, in principle, be carried out in the field with the existing gas cabinets. For instance, under-ventilation can be added by lifting the casing a few centimeters from its base, creating a slot at the bottom. In addition, it is possible to raise the roof edge a few millimeters by unscrewing it and replacing it with longer bolts with thicker spacers (o-rings). The effects of these practical, applicable modifications were examined. This revealed the following:

Recommendations:

The following recommendations emerge from this study:

• Consider incorporating concentration measurements into standard operating procedures by recording them for research and monitoring over longer periods. This can provide interesting insights into how populations of gas pressure regulating stations evolve.
• Consider adapting work instructions so that, when working on gas stations, technicians take a gas concentration measurement in the vent opening before opening the cabinet itself. By recording the measured gas concentration, network operators will gain more insight into actual frequency of occurrence of larger leaks with limited extra effort. In addition, leaks found can then be repaired, reducing the population of stations with leaks.
• The application of hydrogen at gas stations seems to require additional precautions to achieve the same level of safety as for natural gas. For the 4 m³ cabinet station, this difference is more evident than for the ½ m³ cabinet. Based on the precautionary principle, for both the ½ m³ and the 4 m³ cabinet station, additional precautionary measures are sensible. There are several possibilities here, such as further increasing the ventilation area of the gas cabinet, modifying the gas cabinet, placing fencing around at least one meter away from the gas cabinet or more intensive monitoring for leaks than is usual for natural gas stations.
• Further research on HAS cabinets (mini-cabinets) is recommended before HAS cabinets are converted to hydrogen. Until such research is conducted, it is advisable to apply additional precautions when distributing hydrogen.
• It is important that the standards committee of NEN1059 makes a statement for which leakage flow rates ventilation should be an effective measure under which specific circumstances. Leakage openings and practical measurements still seem to be far apart.