Derating a 10 barg Air Receiver Found Thinned by UT
A 20-year-old vertical air receiver at a bottling plant has a shell and heads of SA-516 Gr 70, inside diameter 1200 mm, original shell thickness 8.0 mm, design 10 barg at 60 °C, full radiography (E = 1.0) and a corrosion allowance of 1.5 mm. The compressor is rated 5.5 m³/min FAD with a cut-out pressure of 9.5 barg. A routine UT survey reports a general wall thickness of 7.4 mm, but a 300 × 200 mm patch on the lower shell just above the support skirt averages 4.6 mm, with isolated pits down to 3.9 mm. The existing relief valve is set at 10.0 barg. The plant runs three shifts and cannot stop the compressor for more than 24 hours. As the COREN-registered engineer holding the pressure-equipment register, you are asked either to certify continued operation at 10 barg or to state the reduced duty you will sign for. The allowable stress at design temperature is 138 MPa.
1. Can the vessel stay at its stamped design pressure? Calculate what pressure you would authorise and how the local pitting affects your answer.
Read the worked answer
Ring tension governs, so use the thin-shell formula for the circumferential seam:
- t_required = P·R / (S·E − 0.6P) + CA
- t_required = (1.0 × 600) / (138 × 1.0 − 0.6 × 1.0) + 1.5 = 4.35 + 1.5 = 5.85 mm
Measured 4.6 mm is below 5.85 mm, so 10 barg cannot be supported and the certificate must not be renewed at the stamped pressure. Now invert the formula with the measured thickness, taking the corrosion allowance already consumed:
- t − CA = 3.1 mm
- P = 138 × 1.0 × 3.1 / (600 + 0.6 × 3.1) = 427.8 / 601.9 = 0.711 MPa ≈ 7.1 barg
Round down to 7.0 barg: pressure ratings are not rounded up because the formula is already a lower-bound estimate of a nominal wall.
The 3.9 mm pits are a separate question. Applying the same formula to 3.9 mm gives only about 5.5 barg, but the shell formula is not the correct tool for a local thin area of limited extent - a fitness-for-service assessment of the pitted patch (level 1 then level 2) is required. My position: stamp 7.0 barg based on the 4.6 mm average, conditional on the FFS assessment of the pits accepting 7.0 barg; if it does not, the stamp falls to 5.5 barg. Either way the pit patch gets a close-interval UT grid and a corrosion-monitoring point.
2. What must change on the protection side, and will the existing relief valve still protect the compressor at the new pressure?
Read the worked answer
The relief valve setting is the last line of defence, so it must follow the reduced MAWP or the derating is meaningless:
- Relief valve set at the new MAWP: 7.0 barg, with the spring re-rated/replaced and the valve re-tested and re-sealed. It cannot simply be left at 10.0 barg.
- Compressor cut-out lowered to about 6.5 barg and cut-in to about 5.5 barg, so normal control keeps the machine below the set pressure and the relief valve only sees a control failure.
- No isolating valve between receiver and relief valve; verify the vent discharges to a safe location.
Capacity check at the lower set pressure: the existing orifice is rated 9.0 m³/min at 10 barg. Critical-flow capacity through a fixed orifice scales roughly with absolute inlet pressure for the same gas and temperature:
- 9.0 × (8.0 / 11.0) ≈ 6.5 m³/min against 5.5 m³/min compressor FAD
That is adequate, but the margin has shrunk, so the valve data sheet must be re-issued with the new set pressure and capacity and the sizing checked, not assumed. The register entry, nameplate and inspection interval also change: shorter external inspection interval (6-monthly), internal examination within 12 months, and a written scheme of examination re-dated. Tell the plant that the derated vessel is now a lower-pressure machine with the same leak and brittle-fracture risk - nothing about the derating makes inspection optional.
3. The owner asks you to approve an in-situ weld repair of the thinned patch so the receiver can go back to 10 barg. Give your engineering judgement, including how you would sequence repair versus replacement.
Read the worked answer
I would not approve in-situ weld repair of a thinned patch on a receiver that is in air service and has measurable pitting, for these reasons:
- Weld build-up over a corroded, oil-contaminated surface is a poor metallurgical repair; the correct repair is a patch plate or insert with a full-penetration weld, which needs radiography of the new seam and, in practice, workshop conditions.
- The vessel is 20 years old with general loss from 8.0 to 7.4 mm; a local repair restores one patch but leaves the remaining life question open and may cost a large fraction of a new receiver.
- Hot work on an in-service air receiver requires gas-freeing, cleaning and a permit regime, and it cannot be done credibly inside a 24-hour compressor shutdown with the plant still operating around it.
Sequence I would sign:
- Derate to 7.0 barg now, re-set the relief valve and controls, update the nameplate and register.
- Order a replacement receiver with a proper corrosion allowance, drain leg and internal coating, delivered within 6 months.
- If a repair is still preferred, remove the vessel from service and repair it in a workshop under a qualified procedure with radiography, then re-rate by calculation and hydrotest.
- Regardless of route, find the water source causing the bottom-shell corrosion - a receiver that thins above the skirt is telling you about condensate and drainage, not about steel quality, and the new vessel will do the same if the drain trap is not fixed.