Knowledge / Shipyards and production

Ship piping pressure tests: test boundaries, stored energy and interpretation

Understand hydraulic and pneumatic test energy, temporary-blank loads, elevation effects and the limits of a pressure trace.

On this page

A pressure test produces evidence about a piping system in a defined condition. Its value depends on which components lie inside the boundary, the test medium, the location of pressure measurement and the acceptance question. In a shipyard, a temporary blank or connector weaker than the permanent installation can become the most critical component of the entire test.

Strength, tightness and function are different questions

A proof test examines structural integrity under a specified load. A leak test evaluates leakage or leakage rate under defined conditions. A functional test examines behaviour such as valve movement or protective action. One record does not establish all three. A closed control valve, for example, can leave the downstream section unpressurised while the inlet gauge displays the intended pressure.

The test boundary should be explicit on the piping diagram. Included and excluded equipment, temporary components, pressure tapping points and isolated pockets must correspond to the actual installation. Acceptance pressure, duration and examination method come from the applicable design code and approved plan. There is no single pressure multiplier suitable for every ship piping system. Temporary connections must suit the test pressure and medium to avoid a weak boundary component. A suitable pressure-limiting or relief provision prevents the source from loading the boundary above its intended test pressure.

Choosing a medium also changes stored energy

HSE GS4, fourth edition, relates test hazards to stored energy, volume, pressure and medium, and recommends suitable liquid rather than gas wherever possible. It is UK occupational-safety guidance, not another flag State's piping acceptance rule.

Compressed gas can perform substantial work as it expands. A liquid-filled assembly still stores energy through liquid compression and elastic deformation of pipes and hoses; it is not energy-free. A trapped gas pocket can materially change that behaviour. Two tests reaching the same gauge pressure therefore cannot be assumed to present the same hazard. Planned venting in a liquid test removes compressible gas pockets; choosing liquid alone does not establish that those pockets are absent.

Liquid also introduces weight, corrosion, contamination, freezing and subsequent drying considerations. NASA Glenn's historical pressure-safety document, 2019 revision with its 2024 update gives institution-specific examples of reasons to consider pneumatic testing, including inability to support liquid weight or tolerate residual liquid. These do not grant permission for a shipboard test. The alternative needs its own engineering and safety assessment.

Force on a temporary blank

Assume a hypothetical circular pipe end with an effective diameter of 0.20 m and an internal-to-external pressure difference of 1.0 MPa. Area A = π D² / 4 = 0.03142 m², giving an end force F = Δp A ≈ 31.4 kN. This is the resultant force carried by the blank and its entire restraint path. It does not specify an acceptable blank thickness or bolt selection.

Doubling diameter at the same pressure quadruples area and end force. A connection concept used for a small pipe cannot consequently be scaled up merely by choosing visually larger components. Blank bending, the bolt group, gasket seating and support stiffness need separate assessment. The example calculates force; it does not calculate fragmentation, projectile travel or a safe exclusion distance.

Gauge location changes the interpretation

Under static conditions, pressure at a lower point exceeds that at a higher point by ρ g h. With water density assumed to be 1,000 kg/m³, g = 9.81 m/s² and an elevation difference of 12 m, the difference is 117,720 Pa, approximately 1.18 bar. A 10.0 bar reading at the upper point corresponds to approximately 11.18 bar at the lower point. These are teaching assumptions only.

A system can reach the required pressure at its highest elevation while overstressing a lower-rated component below. Gauge range alone is therefore insufficient: measurement location and the distinction between gauge and absolute pressure matter. During flow, friction and transients add further effects that the static elevation calculation does not cover.

Interpreting the pressure trace

Falling pressure is not necessarily leakage. Gas cooling, volume change, gasket settlement or a change in the measurement chain can contribute. For an ideal gas at fixed volume, absolute P/T is constant; using Celsius or gauge pressure in that relation is incorrect. A pressure trace without temperature information cannot reliably establish a precise leak rate.

Constant pressure does not establish correct coverage either: the pressure source may replenish a loss, or an isolated pocket may remain outside the measurement. Medium, tapping location, instrument identity, temperature conditions and pressure-source state should accompany the record. An abnormal result calls for controlled depressurisation and an approved correction process, rather than tightening a connection while it remains pressurised.

What the completed test means

A successful test does not establish freedom from future fatigue, corrosion or vibration damage. NASA's pressure-testing explanation distinguishes pressure cycling and potential flaw growth from static testing. Space-hardware acceptance criteria are not transferred here to marine pipework.

Returning a system to service requires accounting for temporary blanks, test connections, isolated equipment and drainage or drying status. Zero on one gauge does not establish zero pressure in every isolated volume. The certificate is useful when it remains connected to the actual tested arrangement and final operating configuration. Keeping people outside the hazard area and arranging observation from a safe location belong to test design as much as measurement quality does. Remote monitoring reduces personnel exposure; controlled release from every isolated volume removes stored energy before re-entry. These functions also appear in GS4 test-design guidance.

Sources