Cold box separator cylinder sealing performance testing is a critical quality assurance process for cryogenic air separation units, LNG facilities, and petrochemical plants, verifying that gas-liquid separator vessels housed inside insulated cold boxes remain leak-tight under extreme low-temperature and cyclic pressure conditions. This technical guide covers the testing samples, testing items, helium leak detection methods, applicable standards, and industrial applications associated with cold box separator cylinder sealing performance evaluation.
Technical Overview
A cold box is a thermally insulated enclosure—typically filled with perlite or aerogel insulation—that houses cryogenic process equipment such as plate-fin heat exchangers, distillation columns, pipelines, and separator vessels. Within this system, the separator cylinder (also referred to as a gas-liquid separator vessel or knock-out drum) serves to disengage liquid droplets from vapor streams at operating temperatures as low as -196°C. Because these vessels operate in a deeply cryogenic, thermally cycled, and often high-purity oxygen or hydrocarbon environment, any loss of sealing integrity can lead to product contamination, cryogenic fluid leakage, internal icing of the insulation bed, energy losses, and potentially serious safety incidents.
Sealing performance testing of cold box separator cylinders is therefore performed during manufacturing, pre-commissioning, and periodic turnaround inspections. The testing program combines strength verification (hydraulic pressure testing), tightness verification (pneumatic and pressure decay testing), and high-sensitivity leak detection (helium mass spectrometry), supported by nondestructive examination of welds. Once the cold box is purged with dry nitrogen and filled with perlite insulation, leaks become extremely difficult to locate and repair, making pre-insulation sealing verification absolutely essential.
Testing is generally conducted in accordance with recognized standards frameworks, including GB 150 (Pressure Vessels), GB/T 15087 and related Chinese pressure vessel technical specifications, TSG 21 (Specialized Equipment Safety Technical Regulations for Stationary Pressure Vessels), NB/T 47013 (Nondestructive Testing of Pressure Equipment), ASME BPVC Section V Article 10 (Leak Testing), ASTM E499 (Helium Mass Spectrometer Leak Testing), ISO 20485 (Non-destructive testing — Leak testing — Tracer gas methods), and EN 1779 (Criteria for choosing leak testing methods).
Testing Samples
Aluminum Alloy Separator Cylinders
Separator cylinders fabricated from cryogenic-grade aluminum alloys such as 5083, 5A02, and 5052 are the most common configuration in air separation unit cold boxes. Aluminum offers excellent low-temperature toughness and is compatible with the aluminum plate-fin heat exchangers installed in the same cold box. Test samples include the cylindrical shell sections, dished heads, longitudinal and circumferential weld seams, and internal demister support structures.
Austenitic Stainless Steel Separator Vessels
For LNG, ethylene, and hydrogen liquefaction services, separator cylinders are commonly manufactured from austenitic stainless steels such as S30408 (304) and S31603 (316L), which retain ductility and impact toughness at cryogenic temperatures. Testing samples cover the vessel body, weld joints, and dissimilar-metal transition joints where applicable.
Welded Joints and Nozzle Connections
All Category A, B, C, and D welded joints as defined by pressure vessel codes are subject to examination, including shell longitudinal seams, girth welds, nozzle-to-shell welds, and attachment welds for internal components. These joints represent the primary potential leak paths in the finished vessel.
Flanged Connections, Gaskets, and Penetrations
Sealing surfaces of flanges, instrument ports, level gauge connections, manways, and pipeline penetrations passing through the cold box shell are tested together with their gaskets (spiral wound, flexible graphite, or metal-jacketed types) to confirm complete assembly tightness rather than vessel-only integrity.
Testing Items
Depending on the design specification, service medium, and applicable code, the sealing performance test program for cold box separator cylinders typically includes the following items:
- Hydrostatic (hydraulic) strength pressure test at 1.25 times design pressure to verify structural integrity and gross leak tightness
- Pneumatic tightness test at design pressure with a hold period for pressure decay measurement
- Helium mass spectrometry leak detection on weld seams, flanged joints, and threaded connections, with typical acceptance thresholds in the range of 1×10⁻⁹ to 1×10⁻⁷ Pa·m³/s depending on service class
- Pressure decay (pressure drop) testing to quantify system-level leakage over a defined stabilization and observation interval
- Bubble emission testing using soap solution or film-forming surfactant for accessible exterior joints
- Vacuum box testing for weld seams where only one side is accessible
- Nondestructive examination supporting leak integrity: radiographic testing (RT), ultrasonic testing (UT), penetrant testing (PT), and magnetic particle testing (MT) where material properties permit
- Gasket seating and seal surface verification, including flatness and surface finish inspection of flange faces
- Thermal cycling leak verification for critical service vessels, in which leak testing is repeated after simulated cool-down and warm-up cycles
- Helium background and residual gas baseline measurement prior to formal leak rate measurement
Testing Methods
Helium Mass Spectrometry Leak Detection
Helium leak detection is the preferred high-sensitivity method for cryogenic separator cylinders because helium, as a small inert molecule with low background concentration in air, reliably reveals defects that would leak cryogenic fluids or high-purity gases in service. Two configurations are commonly applied. In the vacuum hood (inside-out or outside-in) method, the vessel is evacuated and connected to a mass spectrometer while helium is sprayed or enveloped around the exterior; any helium entering the vessel is quantified as a total leak rate, making the method suitable for certification-grade acceptance testing. In the sniffer probe method, the vessel is pressurized with a helium-air or helium-nitrogen mixture and a probe is traversed along welds and joints at a controlled speed and standoff distance to localize individual leaks. Calibration is performed against certified reference leaks traceable to national metrology standards, in accordance with ASTM E499, ISO 20485, or NB/T 47013.
Hydrostatic Pressure Testing
The hydraulic strength test is performed by filling the cylinder with clean water (with chloride content controlled for stainless steel vessels, typically below 25 ppm) and pressurizing to the code-specified test pressure, usually 1.25 times design pressure multiplied by the ratio of allowable stresses. The pressure is held for a minimum of 30 minutes while inspectors examine all welds and sealing surfaces for sweating, drips, or visible deformation. This test verifies both structural strength and gross sealing performance.
Pneumatic Pressure Decay Testing
Where hydrostatic testing is impractical—for example, for vessels that must remain completely dry before perlite filling or insulation installation—a pneumatic test using dry nitrogen or clean compressed air is conducted. The cylinder is pressurized to the specified test pressure, stabilized, and monitored over a defined hold period. Pressure, temperature, and atmospheric readings are corrected to determine the net leak-induced pressure drop, which must remain within the specified allowable decay. Because pneumatic testing stores significant compressed-gas energy, safety exclusion zones and staged pressurization increments are mandatory.
Bubble Emission and Vacuum Box Testing
For localized verification, a surfactant solution is applied to pressurized joints and observed for bubble formation, providing a simple and effective pass/fail screen for accessible connections. For welds accessible from only one side, such as closure welds and cold box shell penetrations, a vacuum box is sealed over a soap-filmed weld region and evacuated; continuous bubble formation or film distortion indicates through-wall leakage.
Cryogenic and Thermal Cycling Verification
For high-consequence service, sealing verification may be extended to include cool-down simulation using liquid nitrogen, followed by repeat helium or pressure decay testing at both ambient and low-temperature conditions. This approach detects leaks that appear only under differential thermal contraction between aluminum or stainless steel shells, nozzles, flanges, and gasket materials.
Applications
Sealing performance testing of cold box separator cylinders is applied across the full spectrum of cryogenic process industries:
- Air separation units producing oxygen, nitrogen, and argon, where separator cylinders operate near -180°C to -196°C in the lower column, upper column, and crude argon circuits
- LNG liquefaction trains and boil-off gas handling systems, where cold boxes contain cryogenic separators for natural gas liquids and methane service
- Ethylene plant cold sections, including demethanizers, deethanizers, and cracking gas separation equipment operating at -100°C and below
- Hydrogen liquefaction and purification plants requiring ultra-high leak tightness to prevent flammable gas accumulation within the insulation space
- Helium recovery and refrigeration systems, where even minute leaks cause loss of expensive process helium
- Cryogenic liquid storage tank internals, backup pump suction vessels, and nitrogen generator cold sections
- Manufacturing acceptance, pre-commissioning checks before perlite filling, and periodic turnaround re-verification of in-service cold box equipment
FAQ
Why must sealing testing be completed before perlite insulation filling?
Once perlite is loaded into the cold box, the granular insulation completely surrounds the piping and vessels, absorbing any leaking fluid and making leak localization practically impossible. Repairing a leak after insulation filling requires complete perlite removal, vacuum breaking, and re-purging, resulting in weeks of schedule delay. For this reason, helium leak testing and tightness testing are contractually and procedurally mandated before insulation is introduced.
What leak rate is considered acceptable for a cold box separator cylinder?
Acceptance criteria depend on the service class specified by the designer. For high-purity and cryogenic hydrocarbon or hydrogen service, total leak rates of 1×10⁻⁹ Pa·m³/s (helium equivalent) per joint or 1×10⁻⁶ Pa·m³/s per vessel are commonly specified. General industrial gas service typically permits 1×10⁻⁶ to 1×10⁻⁵ Pa·m³/s. The governing value must always be taken from the vessel design specification and applicable code.
Which standards govern the test procedures?
Strength and tightness testing are governed by GB 150 and TSG 21 or ASME BPVC Section V Article 10; helium leak testing is performed per ASTM E499, ISO 20485, NB/T 47013, or EN 1779; and weld NDT supporting sealing integrity follows NB/T 47013, ISO 17636 (radiographic testing), or ISO 17640 (ultrasonic testing) as applicable.
Can leaks be detected in an operating cold box?
Indirect detection is possible through oxygen analyzer readings in the nitrogen-purged annular space, abnormal cold box pressure, exterior frost formation, or increased product contamination. However, definitive localization requires shutdown, warm-up, and perlite removal, which is why comprehensive pre-insulation testing at accredited third-party laboratories or qualified in-house facilities is the industry-standard preventive practice.
How often should in-service sealing verification be repeated?
In addition to statutory periodic inspection intervals required by pressure vessel regulations, full sealing verification is recommended after any internal modification, weld repair, significant thermal excursion, or at every major plant turnaround, typically every three to five years, and always before re-establishing perlite insulation.