Hydrostatic testing fills a piping system, spool, tank or pressure vessel with water, raises it to a pressure above its design pressure, holds it there and examines every joint for leaks, and Paul Industries performs it on the process, utility and sanitary systems we fabricate, install, modify and repair. The test is the last step before a new or repaired system is turned over: it proves that the welds, flanges, threaded joints and components are tight, and it produces the signed record the owner keeps for the life of the system. This page covers when a test is required, how the test pressure is set under each code, the procedure, water quality, pneumatic and alternative tests, safety, and what to do when a test does not hold.
What is hydrostatic testing?
It is a pressure leak test that uses a liquid, almost always water, as the test medium. Water is nearly incompressible, so a hydrostatic pressure test stores very little energy compared with the same pressure in compressed gas, which is why every ASME piping and vessel code treats it as the default leak test. The test pressure sits above design pressure to give a margin, but the purpose is to find leaks and gross defects, not to prove the design by test. Weld examination, by visual inspection, radiography or ultrasonics as the code requires, happens before the test and is not replaced by it.
The vocabulary varies by region and trade. In the UK and in some specifications the same test is called hydraulic pressure testing; in plants, people speak loosely of pressure testing piping, which covers hydrostatic, pneumatic and in-service leak tests alike. The codes call all of them leak tests, and the specification for a project says which one applies to each system.
When is hydrostatic testing required?
A leak test is required on new piping and vessels before first use, and after repairs or alterations to the pressure boundary when the governing code, the owner or the jurisdiction calls for it. ASME B31.3, B31.1 and B31.9 each require every new piping system to be leak tested before initial operation, and a hydrostatic test is the standard method.
Hydrostatic testing requirements come from three places, and all three apply at once: the construction code sets the minimum test, the owner’s specification often adds to it with longer holds, tighter water quality or witness points, and the jurisdiction governs boilers and pressure vessels. After a repair, the owner’s mechanical integrity program decides whether replaced spools, new branch connections or weld repairs on in-service piping need a test. A boiler hydrostatic test after a pressure-part repair is set by the boiler code and the National Board Inspection Code and is normally witnessed by the jurisdiction’s authorized inspector. Where a new line is tied into an operating system during a shutdown, the final closure weld often cannot be tested with the rest of the line; that case has its own rules, covered below.
Test pressure by code
The required test pressure depends on which code governs the system, and it is always calculated from the design pressure, not from the operating pressure. The table below is a hydrostatic test pressure chart for the codes we work under most; the code text, the project specification and the component ratings always take precedence.
| Code and system | Minimum hydrostatic test pressure | Minimum hold | Pneumatic test, where permitted |
|---|---|---|---|
| ASME B31.3 process piping | 1.5 x design pressure x ratio of allowable stress at test temperature to allowable stress at design temperature (ratio capped at 6.5) | 10 minutes, then examine all joints | Not less than 1.1 x design pressure |
| ASME B31.1 power piping, including steam | 1.5 x design pressure, without exceeding the maximum allowable test pressure of any component not isolated | 10 minutes | Permitted with limits set by the code |
| ASME B31.9 building services piping | 1.5 x design pressure | 10 minutes | Permitted with limits set by the code |
| ASME BPVC Section VIII Division 1 pressure vessels | 1.3 x maximum allowable working pressure x lowest stress ratio | Held, then examined at not less than test pressure divided by 1.3 | 1.1 x MAWP x lowest stress ratio |
| API 650 field-erected atmospheric tanks | Filled with water to the design liquid level rather than pressurized | Per the standard and owner | Roof and bottom seams tested by other methods |
For most plant piping the hydrostatic test pressure ASME B31.3 requires is simply 1.5 times design pressure, because at ambient test temperature the allowable stress ratio is 1.0 for systems designed at moderate temperatures. The ratio matters for hot service: where the allowable stress at design temperature is lower than at test temperature, the test pressure goes up to compensate.
Hydrostatic pressure test calculation
A worked example shows the method. A carbon steel line with a design pressure of 150 psig at ambient temperature tests at 1.5 x 150 = 225 psig. If the same line were designed for a temperature at which its allowable stress fell to, say, 80 percent of the ambient value, the ratio would be 1.25 and the minimum test pressure would be 1.5 x 150 x 1.25, about 281 psig. Two checks follow. First, the test must not stress the pipe beyond yield at test temperature; B31.3 allows the test pressure to be reduced to the maximum that does not. Second, every component in the test boundary must be rated for the pressure it will see.
Elevation changes the answer on tall systems. Water adds about 0.433 psi per foot of height, so a riser 100 feet tall sees roughly 43 psi more at the bottom than at the top. The code minimum applies at the high point, so the hydrotest pressure at the low point is higher, and that higher figure is the one checked against flange, valve and equipment ratings. A system test is also different from the proof test pressure used to qualify an unlisted component design, which is a separate qualification test on a sample component under its own code rules.
Hydrostatic test procedure, step by step
How to hydrostatic test pipe safely and with a result the inspector will sign comes down to a written test package and a fixed sequence. Each package runs through the same steps.
- Define the package. Mark the test boundary on the P&IDs and isometrics, with the test pressure, test medium, gauge locations, vents, drains and isolation points. Confirm that weld examination is complete and accepted and that any post-weld heat treatment is done, because the test comes after heat treatment.
- Isolate what cannot take the pressure. Remove or blind off relief valves, rupture discs, instruments, expansion joints, pumps, control valves and equipment rated below the test pressure. Each break gets a hydro test blind flange sized and rated for the test pressure, tagged so it is removed afterwards. Valves are not used as test boundaries unless their seat rating allows it.
- Flush. Hydrostatic testing and flushing are usually sequenced together: debris left from fabrication is flushed out first so it does not lodge in valve seats or score sealing faces during the test.
- Fill and vent. Fill from the low point and vent every high point until water runs clear of air. A hydro test vent at each high point is installed during fabrication, not drilled afterwards. Trapped air compresses, stores energy, slows pressurization and makes the gauge reading unreliable.
- Pressurize in stages. Raise the pressure gradually, stop at an intermediate step to look for gross leaks, then bring the system to test pressure.
- Hold and examine. B31.3 requires the test pressure to be held for at least 10 minutes before all joints are examined; the examination pressure may then be reduced, but not below design pressure. Owner specifications often require longer holds, and some require several hours on hydronic and utility systems.
- Depressurize and drain. Release pressure slowly at the test rig, open the high-point vents before draining so the system cannot pull a vacuum, drain completely and dry where the material or service requires it.
- Restore and record. Remove blinds, reinstall the isolated components with new gaskets and complete the test record before the system is released.
Joints, including welds, are left uninsulated, unpainted and exposed until the test is complete, so a weep can be seen. Shop-fabricated spools can be tested before they ship, which takes most of the joints out of the field test; see stainless steel piping and process piping.
Equipment and setup
The hydrostatic testing equipment on a typical plant job is straightforward: a hydrostatic test pump sized for the system volume and pressure, a hydro testing manifold with isolation, bleed and gauge connections, at least two calibrated gauges, a relief valve on the rig, and, where the owner wants a continuous record, a chart recorder for hydrostatic testing or a digital data logger. Fill volume is worth calculating before the job, because large lines take a lot of water and a long time to fill and drain; our pipe volume calculator gives it per length and size.
Hydro test gauge selection follows a simple rule: the test pressure should sit near the middle of the gauge scale, never at either end. Gauges with a range of at least 1.5 times the test pressure, and commonly about twice it, are read accurately and are not overranged by a surge. Gauges carry current calibration certificates, and a second gauge at another point catches a faulty one. A hydrostatic test pressure relief valve on the rig, set a little above test pressure, protects against an overrun from the pump and against the pressure rise that comes when trapped water warms; a sealed, liquid-full system climbs quickly in the sun, which is why recorders track temperature as well as pressure.
Water quality for hydrostatic testing
The test water must be clean and compatible with the material, and for stainless steel it must be low in chloride and drained and dried promptly. For carbon steel, clean fresh water is the norm; if water must stay in the system for long, the owner may specify inhibitor or treatment, and that changes how the water is disposed of.
Hydrostatic testing of stainless steel pipe needs the most care. Chlorides cause pitting and, in warm service, stress corrosion cracking in austenitic stainless, so owner specifications commonly cap chloride in the test water, often at 50 ppm and in some plants far lower. Sanitary and high-purity systems are usually tested with RO or deionized water, or with the plant’s own purified water. Stagnant test water left in stainless piping or tanks invites microbiologically influenced corrosion, which can pit through thin wall within weeks to months. Industry guidance is to drain and dry completely as soon as the test is accepted, preferably within a day and certainly within a few days. Partial draining is worse than none, because the water left in low points evaporates and concentrates whatever chloride it contains. On sanitary systems, the pressure test normally comes before final cleaning and pipe passivation.
Temperature matters too. Cold test water raises the risk of brittle fracture in carbon steel; Section VIII recommends a vessel metal temperature at least 30 F above its minimum design metal temperature during the test, and not above 120 F so the vessel can be approached safely. In freezing weather, the system is heated, tested quickly and drained, or tested with a freeze-protected medium the owner approves. Finally, disposal: water containing inhibitor, biocide, oil or fabrication debris cannot simply go to a storm drain, and a hydrostatic test water discharge permit or the sewer authority’s approval may be needed before draining.
Hydrostatic testing vs pneumatic testing
Hydrostatic is the default; a pneumatic test is used only where water cannot be. A pneumatic pressure test uses air or nitrogen, and because gas is compressible it stores far more energy at the same pressure. If something fails, that energy is released as a blast and flying fragments rather than a spray of water.
| Factor | Hydrostatic test | Pneumatic test |
|---|---|---|
| Test medium | Water, or another approved liquid | Air or nitrogen |
| B31.3 minimum test pressure | 1.5 x design pressure, adjusted for temperature | 1.1 x design pressure |
| Stored energy | Low; water is nearly incompressible | High; requires a calculated safe distance |
| Leak detection | Visible drips and weeps, pressure trend | Bubble solution, sound, pressure decay |
| Typical use | Nearly all liquid, steam and utility piping, vessels and spools | Gas and instrument air lines that cannot take water, lined pipe, systems that cannot carry the water weight or must stay dry |
| Main risks | Water weight, freezing, brittle fracture in cold metal, residual water | Stored energy release, brittle fracture, personnel exposure |
The pneumatic test pressure as per ASME B31.3 is not less than 1.1 times design pressure, with a preliminary check at no more than 25 psi, pressure raised gradually in steps so strains can equalize, and a relief device set no higher than the test pressure plus the lesser of 50 psi or 10 percent of the test pressure. The owner normally has to approve a pneumatic test in place of a hydrostatic one, and the test plan includes a stored energy and safe distance calculation. Compressed air and nitrogen distribution is the most common case on our jobs; see compressed air piping.
Sensitive leak tests, in-service tests and closure welds
B31.3 also defines leak tests that are not pressure tests in the usual sense. A sensitive leak test, required for some toxic services and wherever the specification calls for it, uses the gas and bubble method of ASME Section V Article 10, or another method of equal sensitivity, at a pressure of at least the lesser of 15 psi or 25 percent of design pressure. For Category D fluid service, which is nonflammable, nontoxic fluid at no more than 150 psi and between minus 20 F and 366 F, the owner can accept an initial service leak test, in which the system is brought up gradually to operating pressure with its own fluid and examined.
The closure weld that ties a new, already tested line into an operating system is the common exception. When a test is impractical there, the code’s alternative leak test applies: the welds are examined by 100 percent radiography or ultrasonics, joints that were not tested get a sensitive leak test, and the piping is shown to be adequate by flexibility analysis. Planning those welds is part of every tie-in package we prepare for plant shutdowns and turnarounds.
Hydro testing hazards and safety
The hazards are stored energy, flying parts, high-pressure water jets, water weight and vacuum on draining. Even a hydrostatic test can throw a blind, plug, gauge or fitting if it is underrated or partly threaded, and a pinhole leak at test pressure produces a jet that can inject water through skin. Nobody stands in line with blinds, end caps or temporary closures while the system is at pressure, and joints are approached for examination only at the reduced examination pressure.
Water weight is easy to overlook. Steam and gas lines are designed and supported for the weight of gas, so filling them can overload hangers and supports. Spring hangers are locked with their travel stops, and temporary supports are added where needed. Draining without venting can collapse thin-wall tanks and large thin-wall pipe under vacuum.
For a pneumatic test, the exclusion zone for pneumatic pressure test work is calculated from the stored energy, with ASME PCC-2 giving methods for both the stored energy and the safe distance, and it is barricaded with access controlled. Hydrostatic tests are barricaded too, on a smaller scale, with a hydro testing in progress sign at each approach and one person in charge of the pump.
Acceptance, pressure drop and test failure
The hydrostatic pressure test acceptance criteria under the piping codes are simple: at the examination pressure, no joint or component leaks. A gauge reading that moves during the hold is not, by itself, a pass or a fail, because temperature changes in the water move the pressure; the examination and the temperature-corrected trend decide it. When a hydrostatic pressure test failure happens, the cause is usually one of the items below.
| Symptom | Likely cause | What to check |
|---|---|---|
| Pressure falls, no visible leak | Water cooling, air still dissolving or compressing, a valve seat passing into another system, a leak under insulation or below grade | Temperature trend, drains downstream of boundary valves, re-vent high points, exposed joints |
| Pressure rises during hold | Water warming in sun or from a heated area | Temperature trend, relief valve setting, bleed back to test pressure |
| Weep at a flange | Uneven bolt load, damaged or misaligned gasket, face damage | Bolt-up sequence and torque, gasket type, flange alignment |
| Leak at a threaded joint | Short engagement, wrong sealant, galled thread | Remake or replace the fitting |
| Pinhole or crack in a weld | Porosity, lack of fusion, crater crack | Excavate and reweld under the qualified procedure, re-examine, retest |
| Gauges disagree | Calibration, gauge elevation, a blocked connection | Compare against a third calibrated gauge; correct for height |
A slow hydrostatic test pressure drop with no visible leak is the most common argument on site. On a cool evening a full system can lose pressure from temperature alone, so it is checked against the temperature record and repeated before anyone starts cutting.
Piping, vessels, tanks and exchangers
Hydrostatic testing for pipes is organized by test package: one system, or part of one, with one boundary, one test pressure and one record.
Hydrostatic testing for pressure vessels is done by the manufacturer under Section VIII before the vessel is stamped. In the field, a vessel is retested after repairs or alterations as the jurisdiction requires. When the connected piping has a higher test pressure than the vessel, the vessel is isolated and the piping tested separately. Hydrostatic pressure testing of heat exchangers is done one side at a time: the shell side is pressurized with the channel or bonnet removed where possible, so tube-to-tubesheet joints can be watched from the open end, and then the tube side. See heat exchangers.
Hydrostatic testing of tanks depends on the tank type. Shop-built pressure tanks follow their vessel code. A field-erected atmospheric storage tank gets a hydrostatic test API 650 style, by filling it with water to its design liquid level while the shell, bottom and foundation are watched, and API 653 governs testing after major repairs or alterations to existing tanks. See tank fabrication and process tanks.
Hydro testing valves for shell and seat leakage is the valve manufacturer’s job under the valve standard, such as API 598; in a system test valves are normally left open and tested as part of the line. Hydro testing in oil and gas plants follows the same B31.3 rules for plant piping; cross-country pipelines are tested under pipeline codes, and we do not offer pipeline hydrotesting as a separate service line. See oil and gas.
Test records
B31.3 requires a record for each piping system tested, showing the date, the system, the test fluid, the test pressure and the examiner’s certification of the result. Our records add the marked-up test boundary, gauge serial numbers and calibration certificates, the chart or data log, hold times, and any repairs and retests, so the owner can trace every joint back to a passed test. On pharmaceutical, biotech and food systems the test record goes into the turnover package with the weld maps and material certificates and is referenced in installation qualification; see the sanitary piping turnover package and validation and commissioning.
What drives the cost of a test
The cost is driven by preparation more than by the hours at pressure: the number of test packages and boundaries, blinds and temporary closures, removing and reinstalling instruments and relief valves, temporary supports, water supply, treatment and disposal, drying for stainless or dry-service lines, witness scheduling, and retests. Testing spools in the shop before they ship is the most effective way to keep field test time short, and it matters most during a shutdown, when every hour of a test is an hour of outage.
What Paul Industries does
Paul Industries tests the systems we build and repair, from shop spools and skids to field-installed process, steam, utility, chilled water and sanitary piping, tanks and equipment connections. Our hydrostatic testing services cover the test package and procedure, isolation and blinding, filling, venting, pressurizing and holding, examination, draining, drying, restoration and documentation, and pneumatic and sensitive leak tests where the code and the owner allow them. We test as part of our process piping installation work as well as steam piping, chilled water systems and emergency repairs, and we retest repaired systems before they return to service. We do not offer gas cylinder or fire extinguisher testing. Request a quote for testing on a new installation, a repair or a shutdown scope.
Standards referenced: ASME B31.3 · ASME B31.1 · ASME B31.9 · ASME BPVC Section VIII Division 1 · ASME BPVC Section V Article 10 · ASME PCC-2 · ASME B16.5 · API 650 · API 653 · API 598
Frequently asked questions
Is there a difference between hydro test and hydrostatic test?
No. Hydro test, hydrotest and hydrostatic test are the same procedure: a liquid-filled pressure test, almost always with water, held at a set pressure above design while every joint is examined for leaks.
Why is the hydro test pressure 1.5 times design pressure?
The factor gives a margin over the highest pressure the system is designed for, so a joint that would leak in service leaks during the test instead. It is set low enough that, for ordinary systems, the pipe stays well below yield. B31.3, B31.1 and B31.9 all use 1.5 for piping; B31.3 raises it further for systems designed for high temperature.
How long does a hydrostatic test take?
The code hold is short, at least 10 minutes under B31.3, B31.1 and B31.9, but the whole test usually takes a shift or more once isolation, filling, venting, examination, draining and restoration are counted. Large systems, long owner holds and stainless drying requirements extend it.
How much does hydrostatic testing cost?
It depends mostly on preparation: how many test packages, blinds and isolations there are, how many instruments come out, whether temporary supports or treated water are needed, how the water is disposed of and whether the system must be dried. A system designed with test vents, drains and break flanges in the right places costs far less to test than one where they have to be added.
Can hydrostatic testing cause damage?
It can, if it is done badly. Overpressure from an unprotected pump or warming water, testing a component rated below the test pressure, water weight on supports designed for gas, vacuum on draining, freezing, and stagnant water left in stainless steel are the usual causes. A written procedure, a relief valve on the rig and prompt draining prevent them.
Why is hydrostatic testing safer than pneumatic testing?
Water barely compresses, so a water-filled system holds little stored energy and a failure releases a spray rather than a blast. A gas-filled system at the same pressure stores much more energy, and a failure can throw fragments a long way, which is why pneumatic tests need a calculated safe distance.
Is hydrostatic testing the same as pressure testing?
It is one kind of pressure test. Pressure testing also includes pneumatic tests with air or nitrogen, combined hydrostatic-pneumatic tests, sensitive leak tests and initial service leak tests. The governing code and the owner decide which applies.
What is the difference between hydraulic and hydrostatic pressure testing?
In piping work the terms usually mean the same thing: a test using a pressurized liquid. Hydraulic is the more common term in the UK and on some equipment; hydrostatic is the term the ASME codes use.
Hydrostatic test vs leak test: which does the code require?
The codes require a leak test, and the hydrostatic test is the default way to meet it. Pneumatic, sensitive and initial service leak tests are alternatives the code allows in defined cases, usually with the owner's agreement.
Hydro test pressure vs flange rating: which one governs?
Both apply. The system test pressure must not exceed what any component in the boundary can take. ASME B16.5 flanges are shell tested by the manufacturer at 1.5 times their 100 F rating, so where the calculated test pressure would exceed that, the test pressure is reduced, the flange is isolated or the class is raised, as the engineer decides.
How often should hydrostatic testing be done?
Plant piping and vessels are tested when new and after repairs or alterations, not on a calendar. In-service condition is tracked by the owner's inspection program with thickness measurements and visual inspection. Periodic retest intervals apply to products such as gas cylinders and fire extinguishers, which are a different field.
What goes in a hydrostatic test report?
The date, the system and test boundary, the test fluid, the test pressure and hold time, the gauges used with their calibration, the result, and the signature of the examiner and any witness. Repairs and retests are recorded on the same report so the history stays together.
Do stainless steel systems need to be dried after hydrostatic testing?
Yes. Water left standing in stainless steel, especially water with chloride or bacteria, can cause pitting and microbiologically influenced corrosion. Drain completely, blow out low points and dry with clean air or nitrogen, and do it as soon as the test is accepted.
Can a closed valve be used as the test boundary?
Only if the valve's seat is rated for the test pressure and the owner accepts it. Most test boundaries use blind flanges or spectacle blinds, because a passing valve seat makes the test unreliable and can pressurize a system that was never meant to see the test pressure.
Does a tie-in weld made during a shutdown need a pressure test?
Normally the new line is tested before the shutdown, leaving only the closure weld. If testing that weld with the system is impractical, B31.3 allows an alternative: 100 percent radiography or ultrasonic examination of the weld, a sensitive leak test where applicable and a flexibility analysis.
Do you test systems you did not install?
Yes, when we are repairing, modifying or tying into them. We review the existing design data and component ratings, build a test package for the affected part of the system and retest it before it returns to service.
Related: Process Piping · Stainless Steel Piping · Steam Piping · Compressed Air Piping · Chilled Water Systems · Plant Shutdowns & Turnarounds · Request a quote