This article is meant to provide some guidance to mechanical refrigeration operators (e.g. ammonia refrigeration processes) as to when a “pressure test” is required and when a “leak test” is required and what are the specifics that have to be met in order to comply. Not having performed a “pressure test” on field erected piping BEFORE the introduction of anhydrous ammonia seems to be a common finding in audits and regulatory inspections. Here is what ASME B31.5, Refrigeration Piping and Heat Transfer Components says, with some explanations provided:
NOTE: the layout of these requirements have been altered by me to better show the breakdown of these requirements and the emphasis on some items has been added by me
538.4 Field Testing of Refrigerant Piping
538.4.1 General Requirements.
Refrigerant piping ERECTED ON THE PREMISES, except piping components that are factory tested, shall be PRESSURE TESTED and LEAK TESTED AFTER installation and BEFORE operation.
SAFTENG NOTE: This first requirement establishes the minimum requirements for testing field erected anhydrous ammonia refrigerant piping. We see the exception for ammonia piping that is part of a “package” AND the manufacturer of that package provides us with documentation that their “package” was pressure tested. We notice that the piping has TWO (2) tests that are REQUIRED: pressure test AND a leak test. BOTH of these tests are done AFTER the pipe has been installed/erected AND most critically BEFORE we start-up the pipe (e.g. introduce full flow of NH3 to the pipe). Even piping on “refrigeration packages” that were pressure tested by the manufacturer will need to be LEAK tested once it is installed at the facility.
538.4.2 Pressure Test
(a) Piping shall be examined BEFORE pressure is applied to ensure that it is tightly connected. ALL items NOT subject to the pressure test shall be disconnected OR isolated by valves, blanks, plugs, or other suitable means.
SAFTENG NOTE: This is ABSOLUTELY CRITICAL to personnel safety! We are going to be applying a pressure to this piping at 110% of its design pressure – anything that was attached to the pipe during construction for any reason such as hoses and instruments that will NOT be part of the final installation MUST be removed so it does not become a lethal projectile (see members area for pressure test fatality accident reports).
(b) A PRELIMINARY TEST at a gauge pressure of up to 25 psig (170 kPa) may be applied, PRIOR to OTHER TESTING, as a means of locating major leaks.
(c) The TEMPERATURE OF THE PIPING SYSTEM during testing shall be ABOVE the ductile–brittle transition temperature.
CAUTION: Take measures to protect personnel from the potential of rupture of piping components during pneumatic testing of systems.
(d) The means used to furnish the test pressure shall have either:
1) a pressure limiting device or
2) a pressure reducing device AND a pressure relief device AND gage on the outlet side.
The pressure relief device shall be set ABOVE the test pressure, but LOW ENOUGH TO PREVENT PERMANENT DEFORMATION of any of the system components.
(e) The pneumatic test pressure used shall be AT LEAST 110% OF THE DESIGN PRESSURE. The test pressure shall NOT exceed 130% of the design pressure of any component in the system.
(f) For large systems that are NOT COMPLETELY VISIBLE TO THE TESTING OPERATOR, the pressure in the system shall be gradually increased to ONE-HALF (1/2) OF THE TEST PRESSURE, after which the pressure shall be increased in steps of approximately one-tenth (1/10) of the test pressure until the required test pressure has been reached. The test pressure shall be continuously MAINTAINED FOR AT LEAST 10 MIN. It may then be reduced to the leak test pressure per para. 538.4.3(c).
SAFTENG NOTE: This requirement is usually a gap in many pressure testing efforts because the facility does not have a pressure testing procedure. We have seen field erected piping systems that are tested from the machinery room and which include some evaporators OUTSIDE of the machinery room and this requirement was not met. It is CRITICAL that we have a pressure test procedure/program that includes this requirement. We can see that the PRESSURE TEST only has to be held for 10 minutes – yet somewhere/somehow this myth that a pressure test is a 24-hour test is prevalent in this industry. Instead, we need to spend the time RAMPING UP our pressures in our testing per this requirement and once we are at 110% of our design pressure we hold it for 10 minutes. The rationale of this 10 minutes is that when pneumatically testing anything, especially when we are at 110% of the design pressure, there is a lot of energy involved. All we are trying to do with this test it to validate our welds/connections on our field erected piping are structurally sound. ONCE we have validated this with the 10-minute hold, we then REDUCE the pressure down to the DESIGN PRESSURE and then we can SAFELY work around the piping while doing the LEAK TESTS.
(g) Mechanical joints at which blanks or plugs are inserted to blank off or facilitate removal of equipment during the pressure test need not be pressure tested after removal of the blank or plug PROVIDED THE JOINT PASSES A SUBSEQUENT LEAK TEST.
538.4.3 Leak Test.
AFTER THE PRESSURE TEST in para. 538.4.2 is completed, a LEAK TEST shall be performed.
(a) Examination for leaks shall be by the GAS AND BUBBLE FORMATION TESTING as detailed in Article 10, Section V of the ASME Boiler and Pressure Vessel Code, or by other methods of equal sensitivity.
EXCEPTION: Refrigerant vapor detection methods shall be used when refrigerant is used in the test medium. Comply with environmental regulations when venting refrigerants.
(b) Examination shall be made of ALL JOINTS AND CONNECTIONs. The piping system shall show no evidence of leaking.
SAFTENG NOTE: DOCUMENT DOCUMENT DOCUMENT and then document some more that EACH weld, flange, connection, etc. has been bubble tested or vapor detection tested.
(c) The PRESSURE USED FOR LEAK TESTS shall be either the design pressure OR a pressure specified in the engineering design.
SAFTENG NOTE: DOCUMENT DOCUMENT DOCUMENT and then document some more that the pressure in the pipe DURING your leak testing methods was the design pressure or a pressure specified in the engineering design.
538.5 Testing Medium for Refrigerant Piping
(a) A SUITABLE DRY GAS such as nitrogen or air shall be used for pressure testing per para. 538.4.2 (i.e. pressure testing). TOXIC GASES shall NOT be used unless personnel are ADEQUATELY PROTECTED FROM HARMFUL EXPOSURE should leaks occur during testing.
SAFTENG NOTE: It should be stated that ALL personnel must be adequately protected from harmful exposures during this method of testing! This means that if we choose to use NH3 as our testing medium then ALL personnel in the rooms/areas (maybe outside areas could be managed differently) need to be protected from a NH3 leak that could occur during testing. Let me say this about “personnel control” during pressure and leak testing – GET ALL NON-ESSENTIAL PERSONNEL OUT OF THE AREAS! NO ONE should be anywhere near pipe during PRESSURE TESTING and ONLY those workers involved with the actual leak testing need to be in the room/area when using the HHC as your test medium. The minimum PPE for leak testing using NH3 should at least include a full-face APR and I would STRONGLY suggest a supplied-air respirator such as an SCBA (especially of personnel are climbing in pipe runs and could get air lines entangled). Our procedure MUST include a requirement that ONLY enough of the HHC/EHS VAPOR/GAS is introduced to the pipe section being tested to achieve the design pressure! This can be very difficult to calculate/control, thus why so many used a gas from high-pressure cylinders and a metered delivery system so that it is easier to achieve the exact pressure test/leak test pressures.
(b) Oxygen or any medium with a LOWER FLAMMABLE LIMIT (LFL or LEL) BELOW 13% shall NOT be used for pressure testing.
(c) The system refrigerant MAY BE USED for leak testing per para. 538.4.3.
CAUTION: Comply with environmental regulations when testing with refrigerants.
SAFTENG NOTE: see note above about using NH3 as our test medium, but we have come across facilities that actually released over 100 pounds of NH3 during a 24-hour period because they used their NH3 as their test medium and when the test was completed they slowly released the NH3 to the atmosphere. The RQ for NH3 is 100 pounds over a 24-hour period.
(d) Water or water solutions should not be used in testing refrigerant piping.
SAFTENG NOTE: this settles the debate from the petro-chem practices where Hydro-Static pressure testing is the norm/baseline requirement (see ASME B31.3, 345.1 Required Leak Test). In our refrigeration piping RAGAGEP B31.5 we see that using water or water solutions is a NO-NO, so the preferred means is PNEUMATIC TESTING.
So there you have the PRESSURE and LEAK testing requirements for our NH3 refrigeration FIELD ERECTED piping. You may have noticed that the code does not require a written procedure; however, I can assure you that in our world of PSM/RMP a written procedure is a MUST! And if you have different pipe specs for different sides of your refrigeration process (e.g. high side and low side) you NEED to provide specific test pressures in your procedure(s). Lastly, when these tests are done on new or modified piping, it should be the QUALIFIED CONTRACTOR who made the welds/connections that performs these tests AND provides the testing DOCUMENTATION for our files. There is a healthy debate raging in the industry as to when we these requirements apply to existing piping. I strongly recommend that EACH and EVERY time we open our CLOSED LOOP refrigeration process that we PRESSURE TEST and LEAK TEST that portion of the process BEFORE we return NH3 to that portion of the process. I hope this article has helped shed some light on what our field erected NH3 piping should go through BEFORE our start-up.