Pressure testing

Define the question before the pressure

A pressure test stores energy in the product and its temporary closures, connectors and hoses. HSE GS4, fourth edition, ties controls to assembly size, pressure and medium. Obtain the acceptance pressure from the governing specification. HSE, GS4

A proof or strength test examines the pressure boundary under prescribed loading. A leak test seeks leakage under defined conditions. A burst test deliberately seeks failure. Passing one does not establish the others. Oak Ridge National Laboratory distinguishes these objectives in Pressure Testing: Fact and Fiction, dated May 13, 2024. ORNL presentation hosted by National Board

Original ideal-gas illustration, not a test recipe: assume a sealed rigid volume, no leakage, uniform gas temperature and an initial 600 kPa absolute at 300 K. Cooling to 297 K gives:

P₂ = 600 × 297 / 300 = 594 kPa absolute.

The 6 kPa drop occurs without mass loss. Real assemblies may also change volume and exchange heat unevenly. This illustration explains the temperature confounder; it does not establish a compensation algorithm, acceptable pressure drop or hold time. Absolute pressure and kelvin are essential to this calculation. CTS, ideal-gas basis and thermal effects

Specify the applicable construction or inspection code, edition, equipment class and customer requirements before selecting pressure, hold time, medium and acceptance criteria. There is no universal “1.5 times working pressure for 30 minutes” rule covering vessels, piping, hoses, valves, plastic systems and assembled instrumentation. A component's normal working-pressure rating also does not independently authorize an overpressure test of the complete assembly.

Hydrostatic and pneumatic tests

A hydrostatic test uses a liquid, commonly water where compatible. A pneumatic test uses a gas, such as air or nitrogen. PHMSA's pipeline guidance explains that compressed gas stores substantially more energy than compressed liquid at comparable conditions. The preference for liquid where practicable is therefore consequential, while liquid compatibility, drainage and the product's application must still be considered. Pipeline-specific practices in that guidance are not general vessel or machinery requirements. PHMSA, hydrostatic pressure testing

Hydrostatic testing still involves stored strain energy and hazardous pressure-containing components. Gas pockets undermine the assumption of an essentially liquid-filled test. Plan filling, venting and checks for trapped gas, and consider the mass of the filled assembly. Water UK expressly treats air removal, filling and anchoring as part of preparing a polyethylene pipeline test. Water UK WIS 4-01-03, Issue 1, March 2024, section 4

Do not substitute gas for liquid solely for convenience. The governing specification may restrict that substitution and require additional engineering controls. ORNL's presentation gives code-specific examples of circumstances permitting pneumatic alternatives, such as inability to support a liquid-filled system or incompatibility with the medium. Confirm the actual provision applicable to the item. ORNL, pneumatic-test applicability

Establish the test boundary and procedure

Identify what is included, excluded, isolated or temporarily installed. Account for valves, instruments, seals, relief devices, blanks, plugs and test fittings. Define the pressure measurement location, intended pressure history, inspection stage, permitted access and controlled depressurization. Rate temporary equipment for the actual conditions rather than assuming it inherits the product's rating.

HSE calls for a written safe system of work, segregation of the test assembly and maintenance of test connectors. It identifies fragment ejection, expelled fittings, hose whip and release of the pressure medium as failure consequences. Select protective arrangements and remote observation appropriate to the test rather than relying on an operator's distance judgment. HSE GS4, planning and safeguards

The procedure should distinguish maximum test pressure from any pressure specified for close inspection. ORNL's presentation notes this distinction for ASME pressure-vessel testing and the need to depressurize before repairs. Never interpret a required pressure hold as permission to tighten a leaking joint under pressure. ORNL, pressure-test considerations

Choose measurements that answer the question

A pressure gauge shows pressure at its connection; it does not directly measure leak rate. In hydrostatic systems, elevation changes local pressure. Water UK's polyethylene pipeline specification explicitly addresses elevations along the test section. Confirm that high points reach the required pressure while low points remain within permissible limits, using the applicable method's reference location. Water UK WIS 4-01-03, section 4.2.4

For gas pressure-decay testing, Cincinnati Test Systems describes filling, isolating the source, stabilizing and measuring pressure loss over a defined interval. Converting that loss to leak rate requires known volume, accurate pressure measurement and test duration. Temperature changes affect the result, so establish stabilization and the performance of the complete circuit, including fixtures and seals. A leak in the fixture can otherwise be mistaken for a product leak. CTS, pressure-decay testing

Material behaviour also matters. Water UK's WIS 4-01-03 specifically addresses the viscoelastic response of polyethylene pipelines and uses procedures that account for pressure decay and creep. Its analysis must not be replaced by an arbitrary zero-drop criterion or transferred to rigid metallic systems. Water UK, principles and analysis

Leakage detection and bounded product example

Select the observation method for the required leakage limit and medium. Bubble testing can locate gas leaks; tracer-gas methods can address requirements beyond an ordinary pressure-hold observation. Swagelok's North Carolina/East Tennessee facility describes film-solution and immersion bubble tests separately from inboard and outboard helium methods. These are different measurement arrangements with different limitations, not interchangeable meanings of “leak-free.” Swagelok, testing methods

As a named consumable example, Swagelok lists Snoop MS-SNOOP-2OZ liquid leak detector for −2 to 93 °C, while Real Cool Snoop has a different listed temperature range. This bounds product selection; it does not set system test pressure, observation time or allowable leakage. Confirm surface and service compatibility and required post-test cleaning. Swagelok, liquid leak detector listings

Evaluate, restore and record

Compare results with the specified criteria: leakage observations or measured rate, required pressure/time history, and any deformation or functional checks. Preserve interruptions and anomalies rather than describing an incomplete cycle as a pass. A successful pressure test is evidence for its defined conditions; it does not replace other required examinations or predict unlimited service life.

Retain item and test-boundary identification, procedure/code revisions, medium, temperatures, measurement locations, calibrated instrument identities, pressure/time records, inspection results, witnesses and disposition. Record repairs and retests distinctly. Finish with controlled depressurization, required draining/drying, removal of temporary equipment and restoration of the operating configuration. The report should make clear both what was demonstrated and which acceptance requirements were applied.

Related reference

Inspection and quality documents