There are numerous models that can be used to assess the consequence and risk of leaks and releases.
One such model is HYRAM which is publicly available from Sandia and the US DOE.

Leakage/loss depends on the vessel design. Metallic or metallic lined vessels have extremely low permeability and losses through the vessel walls are typically imperceptible. Conversely, Type IV composite vessels which have non-metallic liners are subject to permeation. They are required to meet maximum permeation rates as part of their certification. Fugitive emissions from piping systems can also be considered and is dependent on the specific design and level of maintenance.

Training personnel and equipping them with portable gas detectors to properly identify the gas that is
leaking can play an important role in both troubleshooting and emergency response.

Releases from high pressure hydrogen systems often make a sound. In those cases, sound might be the
easiest way for a person to know there is a hazard. However, leaks can be relatively small and diffuse,
thereby not making much sound, or alternately large and so loud that they can be very difficult to find. In
both cases, it can be hazardous to move into or through an area.

Using tools inside a fume hood that may have a flammable gas mixture should be prohibited. A properly operating hood of the right capacity should keep the mixture of hydrogen in air inside the hood below the Lower Flammability Limit (LFL) of hydrogen further reducing any risk. 

If the use of tools is necessary, the source of hydrogen should be isolated before the work begins even if the concentration of hydrogen is expected to be below the LFL. It is best practice to leak test equipment before introducing hydrogen to minimize the probability of leaks. If spark resistant tools are suitable for a specific task for working with hydrogen systems, use of such tools will lower the probability of producing a spark. 

If the concentration of hydrogen is less than the Lower Flammability Limit (LFL) of 4% in an inert gas, it is unlikely that a leak of this gas mix will form a flammable mixture as it dilutes into air. For example, industry uses ‘forming gas’, a mixture of 4 to 5% H2 in nitrogen, as an oxide reducing agent in materials processing furnaces and soldering operations. This mixture can also be used in conjunction with a hydrogen detector for leak testing gaseous hydrogen equipment.

Sprinkler systems and other fire suppression means are prescribed per building and fire codes to limit fire spread to other materials. In the case of a hydrogen leak and fire, it is best practice to isolate the hydrogen source, and let any residual hydrogen gas burn out. Even if the initial fire is extinguished, additional leaking hydrogen may accumulate and ignite with the potential for an explosion. 

The HSP recommends against the use of glycols for pressure tests due to the difficulty of adequately removing all glycol that might be left in a system after a hydrotest. The HSP recommends a pneumatic test at 110% of the system maximum allowable working pressure (MAWP), which is acceptable by code. Due to an increased danger with pneumatics vs hydrotesting, establish a pressure test zone for personnel in the area. 

Several concerns with glycols are: 1) Freezing in a liquid hydrogen   system that could lead to safety issues such as blockage of lines and instrumentation. There could be dead legs or low spots that cannot be emptied or cleaned easily leaving enough to freeze important components such as small gauge lines, vent stacks, and relief devices. 2) Freezing at cold ambient temperatures or within dispenser piping that chills the dispensed gas can also lead to blockage of lines and instrumentation. 3) Off-spec hydrogen if not well cleaned from the lines, vessels, instrumentation, and dead legs of a system. Glycol can interact with materials such as aluminum and lead to pitting and corrosion.

Another concern for glycol solutions is that they are flammable. A leak generated during pressure testing could generate a spray that is readily ignitable by nearby ignition sources. For automatic sprinkler antifreeze solutions, the requirement is to use UL-listed solutions that have undergone flammability testing. with the pressures used in applicable systems. These pressures are at least an order-of-magnitude lower than what is typically used for hydrogen piping. This safety concern can be mitigated by preventing personnel and ignition sources from being near the piping when tested.

As a general rule, pneumatic testing is preferred over hydrostatic, especially for systems with complex piping geometry, as long as proper precautions are taken for the Pressure-Volume (PV) energy. The HSP recommends using a pneumatic test at 1.1 times the design pressure using a clean, dry, inert gas such as nitrogen. This is a best practice for field testing where cleaning afterward is even more difficult. Helpful resources that describe precautions to take are: 1) UK HSE G4 Safety in pressure testing, says to (a) perform a hazard analysis considering stored energy, blast effects, and missile formation; (b) develop a written procedure; and (c) examine system prior to test. 2) ASME PCC-2, Repair of Pressure Equipment and Piping, says to limit stored energy in any one test loop to 271,000,000 J (200,000,000 ft-lb) (0.07 tons TNT). If the stored energy requirement can’t be met, barriers must be erected or separation distances in excess of 60 m (200 ft) must be used. The HSE document offers broad guidance, while the ASME document provides more specific information about precautions as a function of the stored energy.

The Panel recommends performing a pressure test at 110% of design pressure. This requirement should be applied to all systems regardless of construction type since the intent is to ensure pressure integrity and proper installation. All fitting types have modes of failure during installation. For example, there are numerous examples where compression fittings have had ferrules installed incorrectly, tubing improperly inserted, and have been inadequately tightened. In addition, leak checking and pressure testing should always be done in accordance with the locally adopted piping code.  Examples include ASME B31.3 and the Pressure Equipment Directive. 

 
System pressure relief devices will usually need to be removed for the test and temporarily replaced with higher setpoint devices to protect the system during the pressure test. Also, consider a proportional acting relief device.  Piping systems do not require a pop-acting ASME relief valve that are used for pressure vessels since piping system relief devices are more likely to chatter. Chatter can lead to lower than intended flow rate and damage or failure of the valve to operate correctly.   
For high-pressure systems, pneumatic testing is almost exclusively done given the challenge of removing water from a hydro test from the system after the test. Although one might question the wisdom of pneumatically testing at such high pressures, precautions can be taken to ensure a safe test, such as requiring an exclusion zone during the testing.