Purging can be accomplished in several ways including by using pressure cycles, sweeping, or vacuum. Sweeping is the least reliable but can be effective on simple pipe runs. In most cases, vent systems are open to the atmosphere and the ingress of air from the outlet is likely. However, the vent system should be designed to handle fire or explosion internally. This
generally is not difficult and is a best practice that can also be found in several standards which are incorporated by reference in Codes. 

While air that enters from the vent stack outlet is expected to be within the vent stack, any leaks or openings in the vent system should be repaired such that they don’t act as a venturi and aspirate air while venting. A continuous source of air ingress should be avoided. 

It is not possible to define ignition potential by just velocity without more data (i.e. pressure, materials involved, direction of impact). Due to the multiple methods of developing an ignition source (friction, impact, electrical charge) and the low ignition energy, it is assumed that hydrogen in the air will ignite (between 4 -74%), as it does 30-40% of the time with no known ignition source (see GH2 chart below). Therefore, to try and manage impingement by velocity as an ignition source is not a practical method to assure no ignition.

Other Information:
The ignition energy of hydrogen is .02 millijoules. By definition, a joule is equal to the kinetic
energy of a kilogram mass moving at the speed of one meter per second.

From “Mechanical Sparks as an Ignition Source of Gas and Dust Explosions” from The Italian Association of Chemical Engineering Online:

Mechanical sparks are small particles, which due to the impact between two objects are torn loose from the surface of one of the two colliding objects. The kinetic energy is turned into heat and deformation work. Mechanical spark generation is dependent on the pressure with which the one object is working against the other, the relative speed between the objects, the friction coefficient and the hardness of the materials involved.

Extrapolation of the experimental results using a model it could be shown that incendive hot surfaces can be generated also at relative speeds of < 1 m/s.

Additionally, tests were performed using a file traveling at 1m/s against a metal surface, and the ignition of hydrogen over many concentrations was observed.

No, but it depends on the application. Nearly all vents less than 4” in size are not purged with N2. This is primarily due to: 1) large flows required to dilute hydrogen below the flammable range, 2) the cost of the nitrogen, 3) the potential blockage of the stack when being inserted a vent header/stack serving a liquid hydrogen system, 4) the potential for backpressure (depending on the source) to damage or restrict operation of relief devices, and the lack of incidents with non-purged system.

However, a nitrogen flow can be a means considered for specific systems warm GH2 system as part of a hazard assessment. For example, a nitrogen purge might be appropriate for a large diameter vent header that operates at very low pressure such that it might not be able to be designed for an internal deflagration. 

If a nitrogen purge is to be used on a liquid hydrogen system, then the vented hydrogen should have a means to be warmed above -320 F to prevent liquefaction or freezing of the nitrogen. N2 is not allowed for the purging of LH2 systems per CGAG-5.5.

There is no specific requirement not to vent liquid hydrogen from a vent system. Best practice would be to only vent gas from the top of the vessel to relieve pressure. If liquid must be vented, it should be vaporized first. 

Note: It is very unusual to have LH2 flow from a liquid tank out the vent system, as the vent system is connected to the vapor space on the LH2 tanks and there is a large amount of heat transfer into the fluid leaving the vent stack due to the large temperature difference between the cold GH2 and the environment (between 400 and 525 degrees F, a large multiplier). 

However, there are liquid-venting scenarios that must be considered during upset conditions such as when a road tanker might have rolled over. Liquid can be vented from the gaseous portion of the system so the system should be designed for that possibility. 

At NASA Cape Canaveral, they used a natural gas line connected to the vent stack outlet with a thermal sensor to make sure the pilot was lit. They may also have had a sensor to ensure the H2 fire was lit. In cases where substantial quantities of unused hydrogen is vented and the timing and amount of the flow rate is known and controlled, flaring might be useful. NASA guidelines stipulate that flaring is appropriate for hydrogen vent rates surpassing 0.2 kg/s (~0.44 lb/s). Flare systems themselves must incorporate pilot ignition, flameout warning mechanisms, and a means to purge the vent line, ensuring comprehensive safety measures are maintained throughout the process.

I am not sure which picture you are referring to so I will attempt to answer.

If you are referring to the incident where a fire occurred and the vent system was damaged, then this may have been due to lack of proper supports and incomplete assembly of the test systems. In the past, vent systems were not pressure tested for strength but that is changing. 

If you are talking about small standard H2 cylinders (250 scf or less) then a review of the rupture disc operation is in order. Most individual cylinders have a lead-backed (referred to as fused) rupture disc (CGA type CG4 or CG5) that does not connect to a vent stack. Due to the required mobility of cylinders, a vent stack is not required. The rupture disc is not supposed to vent hydrogen unless the cylinder is engulfed in a fire both melting the lead back and then overpressuring the cylinder/ rupture disc. If engulfed in a fire, the rupture disc opening Is required to maintain cylinder integrity. Although adding hydrogen to a fire is not a perfect solution, it is better than a cylinder explosion and then still adding hydrogen to the fire. 

An interstage regulator relief device not connected to a vent stack. This design is poor and occurs more frequently than we would like. It is typical for a 2-stage regulator to have an interstage safety relief device that is not connected to a vent stack. If the 1st stage of the regulator fails open, the increased pressure can open the interstage relief device, If this is not connected to a vent system (which it normally is not), hydrogen will flow to wherever the relief device is pointed.

There is no standard which specifically specifies the use of a flapper. A properly designed flapper should provide de minimus restriction to vent flow, yet still provides weather protection which allows for a vertical release of the vent stack flow, which is best from a dispersion and radiation perspective. Flappers are extensively used successfully and safely on nearly all liquid hydrogen road tankers as well as a supplemental means to vent large flows from stationary tanks in a vertical direction. However, vertical vent systems with a flapper must be monitored to ensure the flapper closes after operation to ensure no water enters the vent systems. LH2 systems are especially susceptible to this due to water freezing on the vent stack cap and its components and expansion contraction.

All vent stacks/systems should be bonded and grounded to minimize ignition sources. Higher pressure streams from higher velocities have a greater risk of igniting for several reasons, including particle impingement. Adding mesh could create more impact points for particulate, which would increase the potential for ignition, but would not increase the probability of a DDT. Similarly, high flow releases at high pressure can create supersonic flow leading to a shock wave. While this shock wave may be sufficient to ignite the hydrogen, mesh on the outlet will not create the conditions for a DDT.

Mesh must also be designed for substantial flow area to minimize backpressure on the venting system.

Instead of mesh, the outlet of the vent systems can include small diameter stainless steel bar (for example 1/8” – 1/4”), that does not add much turbulence and applied backpressure.

Liquid hydrogen is rarely vented as a liquid. If liquid hydrogen is vented, there should be a means to ensure that it is fully vaporized. The vent systems for LH2 tanks are connected to the vapor space on the tanks to ensure in most instances, this occurs. Most vents from a liquid hydrogen system will vent gaseous hydrogen, but this gas, may still be as cold as -420 F. There are no code requirements for warming the vented hydrogen, and cold gaseous hydrogen is frequently vented safely through the use of a tall enough stack, preferably with a vertical discharge, such that the cold hydrogen sufficiently mixes with the ambient air before it would otherwise reach the ground. The vent systems for the safety of people, must meet the radiation requirements of CGA G-5.5 and API 521.
 

Some practical considerations include:

1. The wind can greatly affect the rate at which hydrogen rises due to the low weight of the hydrogen molecule.
2. Warming to at least -390 F will ensure that the hydrogen is above neutral buoyancy. However, venting hydrogen into the atmosphere at these cold temperatures can still cool the surrounding air and create downdrafts that can push the hydrogen towards the ground if not vented at a high enough elevation.
3. Uninsulated vent system piping below about -300 F can create conditions where oxygen enriched liquid air can form and drip to the ground. While this can be addressed with proper materials of construction under the piping, it can still create personnel hazards, especially if dripping down the sides of a tall vent stack.
4. Venting cold hydrogen below the atmospheric dewpoint will result in a visible water vapor cloud. While this can draw attention to the cloud and can create some confusion as to the extent of flammable mixtures within the cloud, it can be very difficult to warm hydrogen sufficiently to avoid all water vapor formation.