- Pressure safety valves (PSVs) and pressure relief valves (PRVs) are often used interchangeably — but they have distinct mechanical designs, opening behaviours, and code requirements that determine which belongs in your system.
- PSVs are mandatory for gas and vapour service under ASME Section VIII; PRVs are the standard for liquid overpressure protection, with important differences in set pressure, blowdown, and seat leakage.
- Choosing the wrong device leads to chatter, premature seat damage, and — in worst cases — failure to open under genuine overpressure conditions.
The Fundamental Distinction: Pop Action vs Modulating Action
The most important mechanical difference between a pressure safety valve and a pressure relief valve lies in how they open. A pressure safety valve opens with a sharp “pop” — the disc lifts rapidly to its full rated capacity almost instantaneously once set pressure is reached. This pop action is created by a huddling chamber or nozzle geometry that causes the inlet pressure to act on a suddenly enlarged disc area once the valve cracks open, creating a snap-action opening. This behaviour is essential for gas and vapour service, where the compressibility of the medium means a gradual opening would dissipate insufficient energy and the vessel pressure would continue to rise.
A pressure relief valve opens modulating — it lifts proportionally to the overpressure above set pressure. The disc rises just enough to allow the flow needed to balance system pressure. For incompressible liquids, this is entirely appropriate: a small lift allows a large volumetric flow, and the valve modulates around set pressure without dramatic snap action. The difference matters enormously in practice. Apply a modulating PRV to a gas system and you’ll get chatter: the valve opens slightly, gas escapes, pressure drops, the valve reclosed, pressure builds again, and the cycle repeats thousands of times a minute — destroying the seat in hours.
Definitions Under ASME and API Standards
ASME Section VIII, Division 1 defines a pressure relief valve as the generic term covering both types, with specific subtypes distinguished by service. Appendix M and UG-125 through UG-136 contain the requirements. The API RP 520 and API RP 521 documents expand on sizing methodology and installation, and between these documents a reasonably clear vocabulary emerges:
Safety Valve (SV): Pop-action device for gas/vapour service. Full lift achieved within 10% overpressure above set pressure. Mandatory for steam and compressible fluid overpressure protection on ASME-coded vessels.
Relief Valve (RV): Modulating device for liquid service. Disc lifts proportionally to overpressure. Designed so that full rated capacity is achieved at 25% overpressure for liquid service under ASME.
Safety Relief Valve (SRV): A combination device designed to provide pop action on gas/vapour and modulating action on liquid. The SRV is the most commonly specified device in process plant because most systems contain a mixture of phases or vary between liquid and vapour service depending on operating conditions. The SRV can be field-adjusted with a secondary adjusting ring that alters the huddling chamber geometry to shift behaviour between the two modes.
Set Pressure, Overpressure, and Blowdown
Three key pressure definitions govern overpressure device selection and sizing, and engineers must be precise about each:
Set Pressure: The inlet gauge pressure at which the valve begins to open (or “pops” for a safety valve). Must be ≤ MAWP (Maximum Allowable Working Pressure) of the protected vessel. API 520 Part I requires the set pressure to be at or below the MAWP for the first device; a second device can be set up to 105% of MAWP for fire or process upset cases.
Accumulation: The pressure increase above MAWP during valve discharge. ASME allows 10% accumulation for single-valve installations (3% for steam boilers) and 16% for multiple valve installations. This means a valve set at MAWP must be fully open by 110% of MAWP.
Blowdown: The pressure drop below set pressure at which the valve reseats. For spring-loaded safety valves, blowdown is typically 7–10% of set pressure. Too little blowdown causes chatter; too much means the system runs longer below operating pressure before the valve reseats. Pilot-operated safety valves (POSVs) can achieve blowdown of 2–3%, which is highly advantageous on systems that operate close to set pressure.
Pilot-Operated Safety Relief Valves
Pilot-operated safety relief valves (POSRVs) represent a third category that deserves specific attention. The main valve is held closed by process pressure acting on a piston or diaphragm — the higher the inlet pressure, the tighter the seal. A small pilot valve (itself a precision spring-loaded device) senses inlet pressure and controls the main valve. At set pressure, the pilot opens and vents the dome of the main valve, allowing the full inlet pressure to lift the main disc to full capacity almost instantaneously.
The advantages of POSRVs are considerable. Seat tightness up to 98% of set pressure is achievable — critical on systems that regularly operate at 90–95% of MAWP where a conventional spring-loaded valve would leak through seat imperfections. Blowdown can be set independently of spring force. The main valve can be very large (up to NPS 20 and beyond) while the pilot remains compact and replaceable in service. POSRVs are the preferred choice for high-pressure gas storage, compressor protection, and any system where near-to-set-pressure operation is unavoidable.
The disadvantages are also real. POSRVs require the pilot sensing line to remain free of blockage — plugged pilots have caused catastrophic overpressure failures. Backpressure handling requires careful design: variable superimposed backpressure (from a common header) must not exceed the pilot spring setting. Fouling, polymerisation, or freezing in the pilot circuit are common failure modes in dirty or cold service.
Spring-Loaded Valve Construction and Key Components
Understanding the internal construction of a conventional spring-loaded safety/relief valve helps enormously when specifying, testing, and troubleshooting these devices.
The nozzle is the primary pressure-containing inlet orifice. Full-nozzle designs have the nozzle extending through the body to the outlet — better for corrosive service because the body metal is isolated from the process. Semi-nozzle (or conventional) designs have the nozzle seated in the inlet, with the body exposed to process fluid. Full-nozzle is the engineering standard for chemical and corrosive applications.
The disc and disc holder sit against the nozzle seat. The disc material is critical: for steam, stainless steel is standard; for aggressive chemicals, Hastelloy, PTFE-lined, or exotic alloy discs are used. The disc holder incorporates the nozzle ring (lower ring) that adjusts the huddling chamber geometry — turning this ring changes the pop action and blowdown behaviour. Incorrectly adjusted nozzle rings are among the most common causes of field valve problems.
The spring provides the closing force against set pressure. Spring material is almost universally cold-drawn or hot-wound carbon steel for ambient service, 316 SS for corrosive atmospheres, and Inconel for high-temperature applications. Spring rate and free length are factory-set and must not be altered in the field — shimming springs or modifying spring seats to change set pressure is a code violation and a safety hazard.
Conventional vs Balanced Bellows Valves
A standard (conventional) spring-loaded safety valve is sensitive to backpressure in the outlet header. Any backpressure in the discharge line — whether superimposed (constant) or built-up (from flow through the line) — acts against the disc and effectively raises the set pressure. For conventional valves, API 520 limits superimposed backpressure to 10% of set pressure to ensure the valve still operates at the intended set point.
Where backpressure exceeds this limit — common in refinery relief header systems or where multiple PSVs discharge to a common flare header — a balanced bellows valve must be used. The bellows seals around the disc holder, isolating the bonnet spring from the outlet pressure. The net effect is that only the nozzle pressure acts on the disc area, and set pressure remains stable regardless of backpressure (up to about 50% of set pressure, beyond which bellows structural limits are reached).
Balanced bellows valves cost more, require bellows inspection and replacement on a defined interval (the bellows is a fatigue-loaded component), and must have the bonnet vented to atmosphere (a leak path that can become a safety issue if the bellows fails and process fluid enters the bonnet). Specify them only where conventional valves are insufficient.
Sizing Methodology: API 520 Part I
The API 520 Part I sizing methodology is the industry standard for process plant PSVs and PRVs. The fundamental approach is to identify the governing relief scenario, calculate the required relief flow rate at the relevant conditions, and select a device with a sufficient effective orifice area (designated API letter orifices from D through T) to pass that flow at the required inlet pressure with 10% accumulation.
For gas/vapour service, the critical flow equation applies when the ratio of backpressure to inlet pressure is below the critical pressure ratio. Most gas relief scenarios at reasonable set pressures will be critical (choked) flow through the nozzle, which simplifies the calculation. For steam, the Napier equation or the steam tables approach using isentropic enthalpy drops provides the mass flow. For liquid service, the Bernoulli-based equation with a discharge coefficient Kd of 0.65 (for liquid) applies.
Always confirm the selected orifice is not oversized by more than a factor of two compared to the required area — oversized valves are prone to chatter in liquid service and simmer/leakage in gas service. If no single standard orifice falls within an acceptable range, consider parallel valves (staggering set pressures by 5%) or selecting a valve with a custom-sized nozzle.
Installation Requirements and Common Mistakes
The installation of PSVs and PRVs is governed by UG-135 (ASME VIII) and API RP 520 Part II, and errors here can completely negate a correctly sized and selected valve.
Inlet pressure drop: API 520 limits pressure drop in the inlet piping to 3% of set pressure under full relief flow. Exceeding this causes the valve to experience pressure below set pressure at its inlet nozzle while the vessel pressure remains above set — leading to rapid opening and closing (chatter). A dedicated inlet line, direct vessel nozzle connection, and short straight runs are all part of good inlet design.
Discharge piping reaction forces: When a large valve opens, the momentum of the exiting fluid creates significant reaction forces at every bend in the discharge piping. These must be calculated (API 520 Annex C provides the methodology) and the piping adequately braced. Unbraced discharge lines on large relief valves have caused catastrophic piping failures.
Block valves under PSVs: ASME VIII permits a block valve between the vessel and the PSV only if there is a full-area spare PSV that remains in service, and only with specific administrative controls (lock-open requirements, tagged-out procedures). Never install a single block valve under a sole PSV without understanding the code basis — it may invalidate the vessel’s code stamp.
Rupture disc combinations: Installing a rupture disc upstream of a PSV is an accepted ASME practice that prevents seat fouling in corrosive service and reduces fugitive emissions. The combination must be pressure-tested together, the space between disc and PSV must be vented or monitored for disc leakage, and the combined assembly must be derated appropriately for the disc’s burst pressure tolerance.
Testing, Maintenance, and Inspection Intervals
PSVs are arguably the most safety-critical valves in any process plant, and their testing and maintenance regime must reflect this. API RP 576 “Inspection of Pressure-Relieving Devices” provides the framework. Most facilities adopt a risk-based approach where testing intervals are set by the severity of service (corrosive fluids, high fouling tendency, steam) and the consequences of failure.
Typical testing intervals range from 12 months for severe service to 5 years for clean, non-corrosive gas service with low fouling tendency. Each valve should have an individual inspection record tracking opening pressure, seat condition, spring condition, and any dimensional measurements against OEM tolerances. Valves that consistently fail to achieve set pressure within tolerance (±3% for set pressures above 70 bar) must be investigated — spring relaxation, corrosion, or mechanical damage to the nozzle seat are the usual culprits.
In-situ testing using acoustic emission technology or a portable Mokveld-type tester allows set pressure verification without removing the valve — valuable for continuous process units where spare valves are unavailable. However, in-situ testing cannot assess seat condition, disc lift, or internal corrosion and should not replace periodic full strip and inspect.
Summary: Which Device for Which Application
To close with a practical decision framework: if your protected fluid is a gas or vapour under any operating condition, specify a safety valve or safety relief valve with appropriate huddling chamber adjustment for the service. If the fluid is strictly liquid and temperature/composition changes will never create vapour, a relief valve is appropriate. If the system operates near set pressure, consider a pilot-operated device. If backpressure in a common relief header exceeds 10% of set pressure, move to a balanced bellows design or a pilot-operated valve. And if fouling, corrosion, or process contamination is a concern for seat integrity, evaluate a rupture disc/PSV combination.
Getting this decision right at the engineering stage costs nothing. Getting it wrong — and discovering the error during a genuine overpressure event — is a consequence that no engineer, plant operator, or maintenance team wants to face.