Learnings

Outcomes

Over 750 direct purges, or purge related tests have been carried out during this project. The results provide evidence to fill the knowledge gap regarding direct purging performance between air and hydrogen. 

Key messages from this work are: 

1) Hydrogen purges are generally more efficient than Natural Gas purges. The total volume of air-fuel mixture created in a purge involving hydrogen is likely to be less than one involving Natural Gas. 

In tests, purges from air to hydrogen have been consistently more efficient than purges from air to methane. Purges from both fuel gases back to air have a relatively similar performance to each other. The low density of hydrogen did not present any challenges for direct purging operations. This means that less fuel-air mix, and less fuel in total, is released for a hydrogen purge compared to a Natural Gas purge. 

2) Hydrogen purges are generally more flammable than Natural Gas purges. The flammable volume of air-fuel mixtures created inside the pipe during a purge involving hydrogen is likely to be greater than one involving Natural Gas. 

Although less air-fuel mixture is created during a hydrogen purge, the wider flammable range of hydrogen means that the volume of mixture that is flammable inside the pipe is greater for a hydrogen purge than for a Natural Gas purge. 

3) The total volume of flammable air-fuel mixtures generated outside of the pipe during a purge involving hydrogen is likely to be less than one involving Natural Gas. The upper flammable limit does not prevent vented fuel becoming flammable once it mixes with air outside of the pipe. 

Tests have shown that the total volume of hydrogen required to achieve a purge is often less than that required for a methane purge. This results in more fuel being vented to atmosphere for a methane purge than for a hydrogen purge. Once vented, fuel will disperse and dilute to a flammable mixture. The practical lower flammable limit for hydrogen (~10%) is approximately double that for Natural Gas (~5%). The boundary of the practical lower flammable limit contains a volume of flammable mixture for Natural Gas nearly twice that of hydrogen.

Further work should examine the likelihood and consequence of ignition with hydrogen, compared to Natural Gas during purge operations. This will involve examining the relative hazards of direct purging to hydrogen vs. purging to Natural Gas, and the magnitude of risk associated with each. 

The specific hazards during a purge are: 

·       Flammable mixtures created inside the pipe during the purge 

·       Flammable mixtures created at the vent outside the pipe during the purge 

Events that could result in these hazards causing harm are: 

·       Ignition of flammable mixture inside the pipe during the purge operation 

·       Ignition of a flammable mixture outside the pipe at the vent during the purge operation 

·       Ignition of a flammable mixture outside the pipe during the purge operation, that burns back into the pipe 

This project carried out some ignition consequence testing during purges, in support of the safety case for the large diameter tests. These tests need to be repeated in more detail to quantify the magnitude of the consequences of purge ignitions in a range of pipe sizes, similar to the work reported by British Gas in the 1990s. The likelihood of those ignitions occurring then needs to be calculated and used to assess the associated risk. 

As purge efficiency increases, the consequences of an internal ignition will reduce as the proportion of flammable mix volume to total gas volume inside the pipe will reduce. The smaller the proportion of flammable mix, the smaller the resulting overpressure in the system and hence consequence of that ignition. The likelihood of ignition inside a pipe during a purge is low, as there are very few possible ignition sources inside a pipe. Ignition outside of the pipe is more likely, however the consequences of such an ignition are much less as the volume of ignited gas is not constrained. 

The consequence of an ignition burning back from the outside of the pipe to the inside of the pipe is more severe than an ignition that doesn’t burn back. In the workshop it has been demonstrated that burn back can be prevented by ensuring that the vent speed from a purge is greater than the laminar flame speed. This was achieved with a purge nozzle that restricts flow at the end of the vent pipe. The effectiveness of this has been demonstrated in workshop tests, but more work will be needed to show the effectiveness of this through modification of purge tooling to mitigate the risk of burn back. 

We propose that further work should look at tooling and mitigation of ignition events. The shape and ease of ignition of the plume should be examined to understand the comparative risk of venting hydrogen compared to Natural Gas during purges. 

This project has studied purge efficiency with respect to length, diameter, speed and orientation of pipe and also examined a number of features including ‘n’ sections, blockages, elbows, branches and tees. 

In terms of length, in general the efficiency of the purge improves with length; short lengths of pipe, relative to diameter, result in less efficient purges and greater proportion of flammable mixture being created. In tests, purges to hydrogen performed better than methane.

Faster purge speeds resulted in more efficient purges. This was particularly noticeable at larger diameters; smaller diameters did not require fast purge speeds for efficient purges. It was still possible to purge all of the pipes at slow speeds, but efficiency was reduced. In general hydrogen purges were more efficient than methane purges. When purges were pressure-controlled, hydrogen purges were also significantly faster than methane purges resulting in additional improvements to efficiency. 

The 63 mm diameter pipes generally exhibited efficient purges, once the pipe diameter exceeded 90 mm the speed of purge became more important to achieve an efficient purge. In the larger diameters, speed was a more important factor for purging. 

Pipe orientation is also important. As the pipe angle is increased from 0° to 20° the purge efficiency reduces in methane, then as the angle is increased to 90° the efficiency improves, though not to the same base efficiency. The effect is the same for hydrogen, however the peak inefficiency is at 50°. 

Nearly all of the features tested resulted in acceptable purges. No systems were found that were particularly challenging to purge and all systems that were attempted to be purged were done so without leaving pockets of gas remaining. 

When branches were left un-purged to specifically create a pocket of gas left in the system, the gases mixed through diffusion over time, increasing the concentration of the dead branch. This process was similar in both gases but quicker in hydrogen than methane. 

In nearly all of the tests, hydrogen purges were more efficient than methane purges.

Lessons Learnt

The aim of this project was to deliver the work relevant to H100 Fife, outlining specific work instructions. A comparative study was carried out investigating the purging performance of hydrogen and methane on pipe diameters across the range of sizes to be used by SGN in the H100 Fife project. 

The most significant discovery of the project is that the very low density of hydrogen does not make direct purging between air and hydrogen impossible or even difficult. In many cases direct purging a system in like for like conditions is more efficient for hydrogen than for methane. At the time of writing, it is believed that this is due to the higher coefficient of diffusion for hydrogen. 

These findings should provide confidence that direct purging is a viable option for commissioning and decommissioning the networks for H100 Fife.