Leak Testing of Bellows Seal Globe Valves: Standards, Methods, and Acceptance Criteria

Leak testing of bellows seal globe valves is a critical quality assurance process that verifies the integrity of the pressure-bearing shell, the sealing performance of the seat, and the hermetic capability of the metal bellows stem seal. This technical article explains the applicable standards, sample requirements, test items, detection methods, and acceptance criteria commonly applied by accredited third-party laboratories and valve manufacturers. Understanding these procedures helps engineers specify, validate, and maintain zero-leakage valves for hazardous, high-purity, and high-temperature fluid services.

Technical Overview

A bellows seal globe valve is a linear-motion stop valve in which the stem is sealed primarily by a welded metal bellows assembly instead of, or in addition to, conventional packing. The bellows is typically manufactured from single-ply or multi-ply austenitic stainless steel or nickel alloy and is welded to both the stem and the bonnet, creating a friction-free, static barrier that achieves virtually zero stem leakage over thousands of operating cycles. A backup packing set is normally installed above the bellows as a secondary seal in the unlikely event of bellows failure.

Leak testing of these valves verifies three critical sealing boundaries: the pressure-bearing shell (body, bonnet, and flanged or welded joints), the seat (the closure interface between the disc and seat ring), and the bellows stem seal. Because the bellows is the defining feature of the design, its integrity receives particular attention during both factory acceptance testing and independent type testing. Routine testing is generally conducted in accordance with ISO 5208, EN 12266, and GB/T 13927, while fugitive emission qualification and fire type testing may be performed to ISO 15848 and ISO 10497 where the service conditions demand them.

Testing Samples

Routine Production Valves

Every finished valve intended for shipment typically undergoes 100 percent routine testing. Samples are tested in the as-manufactured condition, including all end connections, with tests performed on the completed assembly after final machining, assembly, and setting of the closure member.

Type Test and Design Qualification Samples

Representative samples of new or modified valve designs are submitted for extended qualification programs. These samples undergo mechanical endurance cycling, thermal cycling, and repeated leak measurements to demonstrate long-term bellows and seat performance, as required for fugitive emission qualification under ISO 15848-1.

Bellows Assemblies and Components

Bellows subassemblies are frequently tested at the component level before valve assembly. Each welded bellows can be individually pressurized and examined by helium leak detection, while weld seams are verified by non-destructive examination such as liquid penetrant testing in accordance with ASTM methods.

In-Service and Maintenance Samples

Valves removed from plant systems during turnarounds, or suspected of leakage during operation, may be submitted to accredited third-party laboratories for verification testing, failure diagnosis, fitness-for-service assessment, or post-repair validation.

Testing Items

Depending on the specification, service severity, and applicable standard, a bellows globe valve leak inspection program may include the following items:

• Shell strength test: verifies that the body, bonnet, and pressure-retaining joints withstand a pressure of 1.5 times the rated pressure without deformation or leakage.
• Shell tightness test: confirms no visible leakage through pressure-bearing walls and static joints at the shell test pressure.
• High-pressure seat leakage test: evaluates closure tightness at 1.1 times the rated pressure with liquid or gas media.
• Low-pressure seat leakage test: evaluates closure tightness with air or nitrogen at approximately 0.6 MPa to detect seat defects that may not appear under liquid testing.
• Bellows stem seal leak test: verifies the hermetic integrity of the bellows and its attachment welds, typically using helium tracer gas.
• Backup packing tightness test: confirms the secondary packing seal functions correctly if the bellows is breached.
• Fugitive emission measurement: quantifies leakage at the stem seal and body joints under cycling and thermal conditions per ISO 15848.
• Endurance and cycle life verification: demonstrates that the bellows and seat maintain integrity over a specified number of full-travel operating cycles.
• Fire type test: optional qualification to ISO 10497 demonstrating acceptable leakage during and after fire exposure.

Testing Methods

Hydrostatic Shell Testing

The valve body is pressurized internally with water (with the closure member partially open) to 1.5 times the maximum allowable pressure at 38 °C, following the durations specified in ISO 5208, EN 12266, or GB/T 13927, which typically range from 15 seconds for small sizes up to 120 seconds for larger nominal diameters. The external surfaces are examined for visible leakage, sweating, or permanent deformation. No leakage is permitted through any pressure-retaining wall or joint.

Seat Tightness Testing

For the seat test, the closure member is fully closed and pressure is applied to each side of the seat in turn, with the opposite side open to atmosphere so that any leakage can be collected and measured. The high-pressure seat test is performed at 1.1 times the rated pressure, while the low-pressure test uses air or nitrogen at approximately 6 bar. Gas leakage is commonly measured by bubble collection under water or by flow measurement. Bellows globe valves in critical service are normally specified to Rate A of ISO 5208, meaning no visible leakage is allowed during the test hold period, although less severe applications may permit the defined volumetric rates of lower tightness classes.

Helium Leak Detection for Bellows Seals

Because a bellows valve is specifically designed for zero emission, conventional hydraulic or bubble testing is often not sensitive enough. Helium mass spectrometer leak detection is therefore applied: the bellows cavity or valve is charged with helium or a helium-nitrogen mixture, and a calibrated sniffer probe is passed slowly around the bellows welds, stem exit region, and body joints. Acceptance thresholds for high-duty applications are typically set at 1 × 10⁻⁶ mbar·L/s (equivalent to 1 × 10⁻⁷ Pa·m³/s) or stricter, with semiconductor and vacuum services sometimes requiring sensitivity in the 1 × 10⁻⁹ range. Component-level testing of bare bellows assemblies can also be performed using vacuum hood or bombing techniques for maximum sensitivity.

Fugitive Emission Testing

Type testing to ISO 15848-1 subjects the valve to a combined program of mechanical stem cycling and thermal excursions between defined temperature class limits, with periodic leak measurements at the stem seal and flange joints using a sniffing method referenced against a calibrated leak. The measured concentrations must remain below the limits of the specified tightness class throughout the test in order for the design to be qualified. This testing is essential for valves handling volatile organic compounds, toxic media, or greenhouse gases subject to emission regulations.

Endurance and Life Verification

To confirm bellows service life, sample valves are cycled through full stem travel under differential pressure for a specified number of cycles, commonly between 2,000 and 10,000 or more depending on the purchase specification. Leak tests at the shell, seat, and bellows are repeated at intervals and at the conclusion of cycling to confirm that sealing performance has not degraded. Thermal shock testing between upper and lower temperature limits may be combined with cycling for severe-service qualifications.

Applications

Leak-tested bellows globe valves are specified wherever media loss, toxicity, flammability, or contamination cannot be tolerated. Typical application sectors include:

• Petrochemical and refining: isolation of flammable hydrocarbons, hot oil, and vapor services where fugitive emissions are tightly regulated.
• Chemical processing: handling of toxic, corrosive, or lethal media such as chlorinated and organophosphorus compounds that demand zero stem leakage.
• Heat transfer systems: thermal oil and molten salt circuits where leaked fluid poses serious fire hazards.
• Hydrogen and gas systems: high-pressure hydrogen distribution and fueling infrastructure requiring extremely low leak rates.
• Nuclear power: primary and auxiliary systems with strict containment and radiation protection requirements.
• Pharmaceutical and fine chemical manufacturing: high-purity processes where contamination control is essential.
• Semiconductor fabrication: ultra-high-purity gas and liquid chemical delivery with helium leak rate limits near detection thresholds.
• Steam and vacuum systems: saturated steam isolation and vacuum processing where packing wear is a known failure point.

FAQ

What is the maximum allowable seat leakage for a bellows globe valve?

Most specifications require Rate A of ISO 5208, defined as no visible leakage during the test hold period. For special applications, a quantitative helium leak rate limit such as 1 × 10⁻⁶ mbar·L/s may be imposed on both the seat and the bellows seal.

Why is helium used to test bellows integrity?

Helium is a small, inert, non-flammable molecule that can pass through extremely small leak paths and be detected by a mass spectrometer at concentrations far below what water immersion or bubble methods can reveal. This makes it the reference technique for verifying the hermetic quality of bellows welds.

What test pressures are applied?

The shell strength and tightness test is normally performed at 1.5 times the rated pressure, and the seat leakage test at 1.1 times the rated pressure, with a low-pressure gas seat test at approximately 0.6 MPa. Exact values and durations are defined by ISO 5208, EN 12266, or GB/T 13927 and by the valve class.

Can bellows failure be detected while the valve is in service?

Yes, in most designs the backup packing set above the bellows provides a temporary secondary seal, and periodic fugitive emission monitoring programs can detect developing leaks. If bellows damage is suspected, the valve should be removed and retested in a laboratory.

How often should in-service valves be retested?

Retest intervals depend on the hazard of the media, cycling frequency, temperature severity, and regulatory requirements. Many operators verify valves during scheduled plant turnarounds, and any valve that has been repaired or re-packed should undergo a full shell, seat, and bellows leak test before return to service.

Is fire-safe testing required for bellows globe valves?

Where the valve serves flammable media or the piping class requires it, fire type testing to ISO 10497 may be specified. This demonstrates that the valve maintains acceptable seat and external leakage during and after a defined fire exposure period.