In the industrial fluid control sector, terms like “zero leakage,” “bubble-tight,” and “Class VI shut-off” are frequently thrown around in procurement meetings and engineering datasheets. However, to a valve testing engineer, these terms mean very different things. When specifying an isolation valve, demanding the wrong valve leakage classes can lead to massively inflated project costs, or conversely, a catastrophic failure to isolate a hazardous pipeline.
The confusion almost always centers around the industry’s three most dominant testing standards: API 598, ANSI/FCI 70-2, en ISO 5208. Each standard was developed for a different valve type, a different engineering philosophy, and a different regional context. Yet all three are routinely cited in the same procurement specification, often without a clear understanding of how they interrelate.
A common and costly mistake is applying an isolation valve standard (API 598) to a modulating control valve, or failing to understand the critical difference between FCI 70-2 Class IV and Class VI shut-off capabilities. In this comprehensive engineering guide, we will break down the exact testing procedures of API 598, ANSI/FCI 70-2, and ISO 5208, decode the leakage class limits, and provide a definitive roadmap for specifying the correct standard for your piping system.
The Golden Rule: Isolation vs. Control
Before diving into the allowable leakage rates, you must understand the fundamental division in the valve industry. The standard you choose is dictated entirely by the valve’s primary function:
- Isolation Valves (On/Off): Valves designed to completely stop or allow flow, such as Gate, Globe, Check, Ball, and Butterfly valves. These are governed by API 598 en ISO 5208 [1].
- Control Valves (Modulating): Valves designed to throttle and continuously regulate flow, pressure, or temperature. These are governed by ANSI/FCI 70-2 (equivalent to IEC 60534-4) [2].
Expecting a control valve to achieve the exact same isolation tightness as a dedicated block valve is a fundamental engineering mismatch. Control valves require clearances between moving parts to prevent friction and stiction during continuous modulation, which inherently makes them more prone to microscopic leakage when fully closed.
The table below provides a high-level overview of the three standards before we examine each in detail:
| Functie | API 598 | ANSI/FCI 70-2 | ISO 5208 |
|---|---|---|---|
| Primary Valve Type | Isolation (Gate, Globe, Ball, Butterfly, Check) | Modulerende regelkleppen | All metallic industrial valves |
| Standard Body | American Petroleum Institute (API) | Fluid Controls Institute (FCI) / ANSI | International Organization for Standardization (ISO) |
| Testing Philosophy | Binary pass/fail by seat type and size | Tiered leakage classes (I–VI) | Graded leakage rates (A, AA, B–G) |
| Soft Seat Requirement | Absolute zero leakage (0 drops, 0 bubbles) | Class VI (small bubbles permitted) | Rate A (no visible leakage) |
| Metal Seat Requirement | Size-based drops/bubbles per minute | Class IV (0.01% of rated Cv) | Rate B–G (DN-proportional mm³/s) |
| Test Medium (Primary) | Water; air/gas by agreement | Air, water, or nitrogen | Water or gas (declared per class) |
| Backseat Test | Required if feature exists | Not applicable | Optional |
| Regional Dominance | North America, API-centric projects | Global (control valves) | Europe, international projects |
| EN Alignment | No formal EN route | No formal EN route | Aligned with EN 12266-1 |
API 598: The Isolation Standard
Published by the American Petroleum Institute, API 598 (Valve Inspection and Testing) is the most widely adopted standard for inspecting and testing general-purpose isolation valves in North America and API-centric projects globally [1]. It applies to valves manufactured to API 6D (pipeline valves), API 594 (check valves), API 599 (plug valves), API 602 (steel gate and check valves), API 603 (corrosion-resistant gate valves), API 608 (metal ball valves), and API 609 (butterfly valves), among others.
The API 598 Testing Procedure
API 598 dictates three distinct test types: the shell test (verifying pressure boundary integrity), the seat/closure test (verifying internal sealing), and the backseat test (applicable only to valves with a backseat design, such as gate and globe valves). The seat test involves pressurizing one side of the closed valve and measuring the fluid that escapes past the seat to the other side.
For soft-seated (resilient) valves — those with PTFE, EPDM, or elastomeric seats — API 598 is unforgiving. The allowable leakage rate is zero drops of liquid en zero bubbles of gas for the duration of the test. This is the true definition of “bubble-tight.”
For metal-seated valves (excluding check valves), because metal does not deform to fill microscopic gaps, API 598 allows a specific, permissible leakage rate. The acceptable leakage is calculated based on the nominal pipe size (NPS) [3].
| Valve Size (NPS) | Soft-Seated: Liquid (drops/min) | Metal-Seated: Liquid (drops/min) | Metal-Seated: Gas (bubbles/min) |
|---|---|---|---|
| < 2 (DN 50) | 0 | 0 | 0 |
| 2–6 (DN 50–150) | 0 | 12 | 24 |
| 8–12 (DN 200–300) | 0 | 20 | 40 |
| > 12 (DN 300+) | 0 | 28 | 56 |
Note: 1 mL = 16 drops (liquid); 1 mL = 100 bubbles (gas). For metal-seated check valves, the allowable liquid leakage is 3 cc/min per inch NPS, and gas leakage is 0.042 m³/h per inch NPS [3].
For example, a 6-inch metal-seated gate valve is allowed up to 12 drops of water per minute or 24 bubbles of gas per minute during the low-pressure closure test. The primary test medium is water, with air or gas testing permitted by agreement between the purchaser and manufacturer, subject to safety considerations [3].
ANSI/FCI 70-2: The Control Valve Standard
The Fluid Controls Institute (FCI) developed ANSI/FCI 70-2 specifically for control valves. Because control valves have vastly different internal trim designs (cages, contoured plugs, V-ports), a single “pass/fail” metric like API 598 is impossible to apply [2]. The standard is also published as IEC 60534-4 for international use and was formerly known as ANSI B16.104.
Instead, FCI 70-2 establishes six distinct Leakage Classes (Class I through Class VI). When an engineer specifies a control valve, they must explicitly state which Class the valve needs to meet based on the process requirements.
FCI 70-2 Leakage Classes Explained
The six classes span a wide range of sealing performance, from a theoretical “no test” agreement at Class I to the most stringent bubble-tight shutoff at Class VI. The table below summarizes the key parameters for each class [2] [4]:
| Lekklasse | Max Allowable Leakage | Test Medium | Test Pressure | Typical Application |
|---|---|---|---|---|
| Class I | No test required | — | — | Utility-grade, non-critical service |
| Class II | 0.5% of rated Cv | Air or water, 50–125°F | 45–60 psig | Double-seat or balanced single-seat valves |
| Class III | 0.1% of rated Cv | Air or water, 50–125°F | 45–60 psig | Higher tightness than Class II |
| Klasse IV | 0.01% of rated Cv | Air or water, 50–125°F | 45–60 psig | Metal-to-metal unbalanced single-seat |
| Class V | 0.0005 mL/min per in port dia per psi ΔP | Water, 50–125°F | Max service ΔP | Critical high-pressure, long closure periods |
| Class VI | Per port diameter table (bubbles/min) | Air or N₂, 50–125°F | 50 psig or max rated ΔP | Resilient-seated (PTFE, O-ring) control valves |
For Class VI, the allowable leakage is expressed as an absolute volume based on port diameter. For example, a 6-inch port diameter is permitted a maximum of 4.0 mL/min (27 bubbles/min), while a 3-inch port is limited to 0.9 mL/min (6 bubbles/min) [3].
The Myth of Class VI “Zero Leakage”
It is crucial to note that FCI 70-2 Class VI is technically not “zero leakage.” For a 6-inch valve, Class VI legally permits up to 27 bubbles of air per minute. While practically invisible in a roaring pipeline, it is slightly more lenient than the absolute “zero bubbles” demanded by API 598 for soft-seated isolation valves [2]. This distinction becomes critical when specifying valves for toxic gas service, high-pressure hydrogen, or applications where even microscopic fugitive emissions are unacceptable.
ISO 5208: The International Graded Standard
ISO 5208 (Industrial valves — Pressure testing of metallic valves) is the international standard broadly aligned with the European EN 12266-1. Unlike API 598’s binary pass/fail approach, ISO 5208 relies on a graded system of leakage classes, allowing engineers to contractually tune acceptance criteria against specific operational risks [5]. This flexibility makes it the preferred standard for European and international projects, where the risk profile of a given service can be explicitly encoded into the valve procurement specification.
ISO 5208 Leakage Rates (A through G)
ISO 5208 defines 10 leakage rates: A, AA, B, C, CC, D, E, EE, F, and G, with Rate A being the most stringent [5]. A key feature of the ISO 5208 system is that allowable leakage for Rates AA through G scales linearly with the nominal valve size (DN), expressed in mm³/s for liquids. This DN-proportional approach is more nuanced than API 598’s size-banded approach and allows for more precise specification across a wide range of valve sizes.
| ISO 5208 Rate | Liquid Leakage (mm³/s) | Gas Leakage (mm³/s) | Closest API 598 Equivalent | API 6D Cross-Reference |
|---|---|---|---|---|
| Rate A | No visible leakage | — | Soft-seated (zero leakage) | Soft-seated valves |
| Rate AA | 0.006 × DN | 0.18 × DN | — | — |
| Rate B | 0.01 × DN | 0.3 × DN | Metal-seated NPS < 2 | — |
| Rate C | 0.03 × DN | 3 × DN | Metal-seated NPS 2–6 | Metal-seated non-check valves |
| Rate D | 0.1 × DN | 30 × DN | Metal-seated NPS 8–12 | — |
| Rate E | 0.3 × DN | 300 × DN | — | API 6A metal sealing (hydrostatic) |
| Rate G | 2 × DN | 6000 × DN | Metal-seated check valves | Metal-seated check valves |
Note: Rates CC and EE are intermediate classes between C/D and E/F respectively. Rates F and G represent the most permissive classes, typically used for large-bore check valves [5] [6].
ISO 5208 and EN 12266-1: The European Connection
ISO 5208 was developed with explicit consideration of both EN 12266-1 (the European testing framework) and API 598. The 2015 edition of ISO 5208 added six new leakage levels (AA, CC, E, EE, F, and G) and provided formal cross-reference tables to both API 598 and EN 12266-1 [5]. EN 12266-1 itself meets the requirements of ISO 5208 for seal classification but lacks the AA, CC, and EE ratings. For European projects, specifying ISO 5208 provides broader coverage and is considered the more comprehensive choice.
How to Select the Right Standard: A Practical Decision Framework
For piping engineers and procurement teams, the choice of leakage standard is not merely academic — it has direct consequences for valve cost, lead time, and long-term reliability. The following framework provides a structured approach to standard selection.
Step 1: Identify the valve function. Is the valve an on/off isolation device or a modulating control element? If it is a control valve, specify ANSI/FCI 70-2 and proceed to select the appropriate class. If it is an isolation valve, proceed to Step 2.
Step 2: Identify the project’s regional and contractual context. If the project is governed by API standards (common in North America, the Middle East, and Asia-Pacific), specify API 598. If the project is governed by EN or ISO standards (common in Europe and international EPC contracts), specify ISO 5208 with the appropriate Rate.
Step 3: Assess service severity. For high-containment services (toxic gases, hydrogen, high-pressure steam), specify the tightest applicable class: API 598 zero leakage or ISO 5208 Rate A. For moderate services, ISO 5208 Rates C–D may be acceptable after a formal risk review. Always document the decision in the Inspection and Test Plan (ITP).
The table below summarizes recommended standards by application type:
| Application | Recommended Standard | Leakage Class/Rate | Notes |
|---|---|---|---|
| General on/off isolation (API project) | API 598 | Zero (soft seat) / Size-based (metal seat) | Default for North American projects |
| General on/off isolation (ISO/EN project) | ISO 5208 | Rate A (soft seat) / Rate C (metal seat) | Aligned with EN 12266-1 |
| Modulating control valve | ANSI/FCI 70-2 | Class IV (metal seat) / Class VI (soft seat) | Do not apply API 598 to control valves |
| ESD valve (baseline) | ANSI/FCI 70-2 | Class IV minimum (per API 553) | Upgrade to Class V/VI for critical services |
| ESD valve (critical/toxic service) | API 598 or ISO 5208 | Zero leakage / Rate A | Fulfilled by high-performance TOVs |
| Pipeline valve (API 6D) | ISO 5208 (via API 6D) | Rate A (soft seat) / Rate C (metal seat) | API 6D references ISO 5208 rates |
| Offshore/subsea isolation | API 598 + ISO 5208 | Zero leakage / Rate A | Dual compliance common in offshore specs |
Specifying the Right Standard for Severe Service
When selecting valves for Emergency Shut-Down (ESD) applications or severe service isolation, the stakes are exceptionally high. According to API 553, an ESD valve must perform at ANSI/FCI 70-2 Level IV as a baseline, but critical applications often mandate Class V or Class VI [7].
Historically, achieving Class VI bubble-tight shutoff required soft elastomeric seats (like PTFE), which are highly vulnerable to high temperatures and fire. This forced engineers to compromise between tight sealing and thermal resilience.
Today, advanced engineering in the Drievoudige offset vlinderklep voor ultrahoge druk allows for metal-to-metal seating that achieves true bidirectional zero leakage in accordance with API 598 [8]. Because the seal is torque-energized and friction-free, this zero-leakage capability is maintained even after thousands of cycles.
Bij Carter Valves zijn onze Six-Eccentric Hexa Butterfly Valves utilize a patented frictionless cam geometry. This not only ensures API 598 zero leakage and ISO 5208 Rate A compliance but also reduces operating torque by over 30% compared to industry-standard triple offset butterfly valves, providing unmatched reliability for critical isolation. For engineers seeking a valve that eliminates the traditional trade-off between tight sealing and high-temperature performance, the Hexa platform represents a significant advancement in isolation valve engineering.
Common Specification Pitfalls to Avoid
Even experienced engineers can fall into costly traps when specifying valve leakage standards. The following pitfalls are among the most frequently encountered in procurement and vendor inspection.
Applying API 598 to a control valve. API 598 is designed for on/off isolation valves. Applying it to a modulating control valve is technically incorrect and can result in the valve being rejected during factory acceptance testing, even though it meets its intended FCI 70-2 class. Always route control valves to ANSI/FCI 70-2 and isolation valves to API 598 or ISO 5208.
Demanding “Class VI” on a high-temperature metal valve. FCI 70-2 Class VI is specifically designed for soft-seated (resilient) control valves. Specifying Class VI on a metal-seated valve operating above 260°C is technically infeasible with conventional designs, as PTFE and elastomeric seats cannot survive at those temperatures. The solution is either a precision metal-seated Class V valve or a Triple Offset Butterfly Valve with a metal/graphite laminated seal.
Edition drift. The allowable leakage values in API 598 have changed between editions. Always state the specific edition in the purchase order and inspection test plan. Failure to do so can lead to disputes between the purchaser and manufacturer about which acceptance criteria apply.
Confusing ISO 5208 Rate A with “zero leakage.” Rate A means “no visually detectable leakage,” which is functionally equivalent to API 598 soft-seat requirements. However, it is not an absolute zero — it is a visual inspection criterion. For applications requiring quantified near-zero leakage, Rate AA or a contractually specified measurement method should be used.
For further guidance on fugitive emissions and leakage standards in the context of environmental compliance, see our detailed article on Fugitive Emission Standards for Butterfly Valves. For ESD-specific leakage requirements, refer to our engineering guide on ESD Valve Selection: Leakage Class and Stroke Time.
Frequently Asked Questions (FAQ)
1. Is API 598 stricter than ISO 5208 for seat leakage?
API 598 is typically a go/no-go standard based on seat type. For soft seats, API 598 demands zero leakage. ISO 5208 is graded; its tightest class (Rate A) is equivalent to API 598’s zero leakage requirement.
2. Which standard covers “Class VI” tight shutoff?
ANSI/FCI 70-2 covers Class VI. It is specifically for control valves. You should not use API 598 or ISO 5208 to specify control-valve leakage classes.
3. Can a metal-seated valve achieve zero leakage?
Traditional metal-seated gate or ball valves typically achieve FCI 70-2 Class IV or V. However, advanced Triple Offset Butterfly Valves (TOVs) with laminated metal/graphite seals can achieve true API 598 zero leakage.
4. What is the difference between FCI 70-2 Class VI and API 598 soft seat tests?
API 598 demands absolute zero bubbles of gas for soft-seated isolation valves. FCI 70-2 Class VI, designed for control valves, permits a very small, quantifiable number of bubbles per minute based on the valve’s port diameter.
5. How do I specify leakage for an Emergency Shut-Down (ESD) valve?
ESD valves are isolation valves that must act quickly and seal tightly. While API 553 suggests a minimum of FCI 70-2 Class IV, critical ESD applications in refining and offshore platforms typically specify API 598 zero leakage or ISO 5208 Rate A, often fulfilled by high-performance TOVs.
6. Where does EN 12266-1 fit in?
EN 12266-1 is the European testing framework that is broadly aligned with ISO 5208. ISO 5208 Rates A through G correspond directly to values in EN 12266-1.
7. Do I need to perform a backseat test?
Under API 598, a backseat test is only required if the valve design actually includes a backseat feature (common in gate and globe valves). ISO 5208 does not mandate a backseat test.
Referenties
[1] NTIA. “API 598 Vs ISO 5208: Valve Testing Acceptance Guide.” Norwegian Technical Inspection Academy.
[2] JH Valve. “Valve Leakage Standards: API 598 vs FCI 70-2 (Class IV vs VI) Explained.” Janhen Valve.
[3] Wermac. “Leakage of Valves – Testing API 598, ANSI FCI 70-2, MSS-SP-61 and ISO standard 5208.” The World of Piping.
[4] SVF Flow Controls. “Complete Guide to Valve Leakage Classes (Class I–VI).” SVF Learning Hub.
[5] Valve Specifications. “Valve Leakage Class, ISO 5208, API 598, EN 12266-1, ANSI/FCI 70-2.” Relia Valve.
[6] Piping World. “Valve Seat Leakage Rates Comparison: API 598, ISO 5208, MSS SP-61 Standards.” The Piping Engineering Resource.
[7] Carter Valves. “ESD Valve Selection: Leakage Class & Stroke Time.” Carter Valves Blog.
[8] Carter Valves. “Drievoudige offset vlinderklep vs kogelkraan: Wat is de juiste keuze voor isolatie bij hogedruk toepassingen?” Carter Valves Blog.