- A correctly sized and set pressure relief valve is the last line of defence against catastrophic overpressure — getting the sizing basis wrong, the set pressure too high, or the back pressure too great can render the device completely ineffective when you need it most.
- API 520 and API 521 together form the definitive engineering basis for relief valve sizing and relief system design in the petroleum and petrochemical industries — understanding the distinction between the two standards is essential for proper application.
- Leaking, chattering, and simmer are the three most common operational problems with pressure relief valves, and all three are preventable with correct specification, installation, and maintenance — not tolerable facts of plant life.
The Critical Role of the Pressure Relief Valve
The pressure relief valve (PRV) — also called a safety relief valve (SRV) or pressure safety valve (PSV) — occupies a unique position in process plant design. Unlike every other valve in the system, the PRV is never meant to be operated as part of normal plant control. It exists solely to protect vessels, pipework, and equipment from overpressure events that, without its intervention, would result in structural failure, explosion, fire, or toxic release.
This special status means that PRV sizing and specification are governed by a separate and more demanding set of engineering requirements than control or isolation valves. The PRV must open reliably at its set pressure, flow its rated capacity without exceeding the maximum allowable accumulated pressure (MAAP), and reclose tightly when normal pressure is restored. Any failure in any one of these functions can have catastrophic consequences.
The relevant engineering standards for PRV design and application in the process industries include API 520 (Sizing, Selection, and Installation of Pressure-Relieving Devices), API 521 (Pressure-Relieving and Depressuring Systems), and ASME Section VIII Division 1 and Division 2 for pressure vessel design. These standards are complementary: ASME defines the vessel design requirements including the maximum allowable working pressure (MAWP); API 521 defines the scenarios that create overpressure demand; and API 520 provides the methodology to size the PRV for those demands.
Understanding the Overpressure Scenarios: API 521
Before you can size a PRV, you need to know what you’re sizing it for. API 521 defines the credible overpressure scenarios that must be considered for each protected vessel or system, and it’s a more involved analysis than most non-specialists realise. The standard categories include: blocked outlet (the most common case, where downstream flow is interrupted while inlet flow continues), fire case (the worst case for many vessels, where external fire evaporates liquid and generates massive vapour flow), cooling water failure, heat exchanger tube rupture, control valve failure (open or shut, depending on configuration), and multiple-case accumulation where several scenarios occur simultaneously.
Each of these scenarios produces a different required relief rate, and the PRV must be sized for the governing case — the one that requires the highest relief capacity. In practice, the fire case is often governing for vessels containing volatile liquids at moderate pressure, while blocked outlet governs for gas-phase systems and high-pressure services where the fire heat input is not the limiting factor.
The fire case deserves special attention because it involves a transient: as the fire continues, the liquid level drops, the vapour space grows, and the relief requirement changes over time. API 521 provides simplified fire heat input correlations based on wetted surface area, and the resulting vapour generation rate at the relieving pressure and temperature must be calculated using the actual fluid properties at relieving conditions, not at normal operating conditions.
API 520 Part I: Sizing the PRV
API 520 Part I provides the sizing equations for PRVs in gas, vapour, steam, and liquid service. The basic approach is to calculate the required orifice area that will pass the required relief flow rate at the relieving pressure (which is the set pressure plus the allowed accumulation, typically 10% of MAWP for single PRV installations and 16% for fire case or multiple PRV installations).
For gas and vapour service, the critical flow equation is: A = W / (C × Kd × P1 × Kb × Kc) × √(T × Z / M), where A is the required orifice area, W is the required relief flow rate, C is a function of the isentropic exponent, Kd is the discharge coefficient (typically 0.975 for vapour-service PRVs, or as certified by the manufacturer), P1 is the relieving pressure in absolute terms, Kb is the back pressure correction factor, Kc is the combination correction factor for installations with a rupture disc in series, T is the relieving temperature, Z is the compressibility factor, and M is the molecular weight.
The back pressure correction factor Kb is one of the most commonly overlooked parameters in PRV sizing. Back pressure — the pressure at the PRV outlet — reduces the differential pressure across the valve and can significantly reduce its flow capacity. For conventional spring-loaded PRVs, back pressure above about 10% of set pressure requires application of the Kb correction factor, which can reduce the effective capacity by 20–40% at back pressures of 30–40% of set pressure. Specifying a balanced bellows PRV or a pilot-operated PRV eliminates the back pressure effect on set pressure (though not necessarily on flow capacity), and is the correct choice for systems with significant back pressure.
Set Pressure Selection and MAWP
The set pressure of a PRV must be at or below the MAWP of the protected equipment. In practice, it’s always good engineering to include a margin between the set pressure and the normal operating pressure — if the operating pressure is routinely close to the set pressure, the PRV will simmer (partially lift and leak) under normal operation, which is both a process problem and a seat damage problem.
The industry convention is that the set pressure should be at least 10% above the maximum normal operating pressure. For a system with a MAWP of 100 barg, a normal operating pressure of 70 barg, and a PRV set at 90 barg (10% above operating, 10% below MAWP), there’s a comfortable margin in both directions. For systems where operating pressure is close to MAWP — common in optimised process designs where the vessels are sized as tightly as possible — the margin between operating and set pressure must be carefully managed to prevent nuisance lifting without compromising overpressure protection.
Multiple PRV installations — where two or more PRVs protect the same vessel — are used when the required relief capacity exceeds the maximum available for a single PRV orifice, or when regulatory requirements mandate redundancy. In a multiple PRV installation, the first PRV is set at the MAWP and the second is set at 105% of MAWP. Both valves may be open simultaneously during large overpressure events (fire case), giving a combined capacity at the allowed accumulation pressure of 116% of MAWP.
Conventional, Balanced Bellows, and Pilot-Operated PRVs
There are three main types of spring-loaded PRV for process service, each with distinct characteristics. Conventional PRVs are the simplest and most reliable — a spring-loaded disc held against a nozzle by the spring force alone. They are sensitive to back pressure: variable back pressure changes the effective spring pre-load and shifts the set pressure. They’re suitable where back pressure is low and relatively constant (typically no more than 10% of set pressure).
Balanced bellows PRVs add a bellows assembly that compensates for back pressure on the disc. The bellows connects the back face of the disc to atmosphere (via the bonnet vent), so that back pressure acts on both sides of the disc equally — it pushes down on the disc from the outlet side, and the bellows transmits the same pressure upward from the bonnet vent side, giving net cancellation of the back pressure effect on set pressure. Balanced bellows PRVs are specified when back pressure exceeds 10% of set pressure, when back pressure is variable, or when a closed outlet header system is used.
Pilot-operated PRVs (POPRVs) use a small pilot valve to sense inlet pressure and control the opening and closing of a larger main valve. The pilot uses the process fluid pressure itself as the actuating force — when pressure rises to set point, the pilot releases the pressure loading on the main disc, allowing the process pressure to pop the disc open against a relatively small spring. POPRVs have several advantages: they open on-ratio (opening sharply only at set pressure rather than simmering), they can be opened to 100% at very low overpressure, and they are not affected by back pressure up to the inlet pressure. They are the preferred type for large-bore installations and for services where tight shut-off after relieving is critical. Their limitation is complexity — pilot-operated valves have more components, more potential failure modes, and require more careful maintenance than conventional spring-loaded types.
Chatter, Simmer, and Seat Leakage: The Three Operational Enemies
Chatter is the rapid, cyclic opening and closing of a PRV that occurs when the valve is sized too large for the relieving requirement. As the PRV opens, it relieves pressure quickly, the process pressure drops below the reseating pressure (set pressure minus blowdown), and the valve snaps shut. The pressure then builds back up, the valve opens again, and the cycle repeats — sometimes at frequencies of several times per second. Each opening-closing cycle damages the seat and disc, and a valve that has experienced significant chatter will have degraded sealing performance and may develop a leak. The solution is correct sizing — a PRV should be sized for the actual relief requirement, not for the maximum possible flow from the system. An oversized PRV installed with the best intentions is a liability.
Simmer is the partial lifting and leaking of a PRV that occurs when operating pressure approaches the set pressure too closely, or when the set pressure drifts down over time due to spring relaxation. A PRV that simmers continuously will eventually develop a permanent seat leak as the leaking fluid erodes and damages the seating surfaces. Maintaining adequate margin between operating and set pressure, and implementing a regular PRV bench testing and reseating programme, are the standard preventive measures.
Seat leakage between relieving events — the valve leaking when it should be fully closed — is often the first indication of seat damage. The API 527 standard defines acceptable seat leakage rates for PRVs before and after testing. For metal-seated PRVs, some leakage is permitted; for soft-seated designs, bubble-tight shut-off is expected. A PRV showing seat leakage above the API 527 limits must be removed from service for repair or replacement, as the leakage will typically worsen over time and the valve may not open at its calibrated set pressure if the seat is damaged.
Testing, Certification and Maintenance
PRVs must be bench tested on a regular cycle — typically annually or every three years depending on service severity and regulatory requirements. The bench test verifies the set pressure (the pressure at which the valve opens), the blowdown (the pressure drop below set at which the valve reseats), and the seat leakage before and after lifting. Test results must be documented, and valves that fail to meet the tolerances defined in API 527 must be reconditioned before returning to service.
In-situ testing using hydraulic testing equipment is possible for some PRV installations where removal from service is impractical, but bench testing remains the most reliable method for verifying all performance parameters. Any PRV that has operated (has opened under pressure) must be assumed to have damaged seats and should be inspected and reconditioned before being returned to service without bench testing — a PRV that lifted under process conditions may have experienced seat damage from the process fluid, debris, or thermal shock that affects its set pressure and reseating behaviour.