Why Flanged Joints Keep Leaking

On a working pipe rack, the bolted flange joint is the only intentional discontinuity in a continuous pressure envelope. Every other part of the line — the pipe, the elbow, the weld, the industrial valves — is dimensioned and inspected to the same standard. The bolted joint is also the only part that has to be re-made every time the system is opened, and it is the part most often assembled by the least-experienced crew on site.

In thirty-plus years of supplying pipe, tube, and flanged components to refineries, chemical plants, power stations, and shipyards, the same five failure patterns turn up in incident reports over and over. They are not exotic. They are not new. They are the failures you can engineer out of the joint with a small amount of discipline on the procurement side and on the boltup side. The rest of this article walks through each one, the spec mistake behind it, and the corrective call-out that prevents the next one.

Failure 1: RF Flange Paired with a Non-Ring Gasket

This is the most common new-build failure. A raised-face (RF) flange is supplied from one vendor, a flat full-face gasket is pulled from inventory, and the joint is assembled with no one noticing that the gasket covers the bolt circle instead of seating inside it. The first pressure cycle blows the gasket, and the line is offline before the hydrotest is finished.

The fix is a spec change, not a procedure change. Every RF flange on the datasheet must be paired with a ring gasket sized to the inner bolt circle. For ASME B16.5 pipe flanges, the call-out is ASME B16.21 for non-metallic ring gaskets and ASME B16.20 for spiral-wound and ring-joint types. For flat-face (FF) cast-iron flanges — common in utility water and HVAC — the spec stays as a full-face gasket, but the call-out must be explicit so the receiving inspector can verify the difference before the joint is made.

Failure 2: Carbon-Steel Stud on a Cu-Ni or Stainless Flange

On a marine cooling line, a desalination package, or a chemical-service line, the flange is often copper nickel flanges or austenitic stainless — but the stud bolt and nut were ordered from a standard fastener catalogue, which means they are carbon steel. The joint assembles cleanly and passes hydrotest. Eighteen months later, the stud is rusted through at the nut face, the nut is seized to the stud, and the joint cannot be re-made for the next turnaround.

This is a galvanic problem, and the solution is a matched-material specification. On a Cu-Ni flange in seawater service, the standard call-out is ASTM A193 B8M studs (316 stainless) with ASTM A194 grade 8M nuts, or Monel K-500 / Monel 400 for the higher-end applications. On an austenitic stainless steel pipe line, B8 / B8M in the same alloy family is the correct pairing. The principle is simple: the stud and the flange must be electrically compatible in the actual service electrolyte, or the stud will always be the sacrificial component.

Failure 3: Wrong Gasket for the Service

A compressed non-asbestos fibre gasket is the cheapest option in the gasket catalogue, and it gets ordered for almost every low-pressure utility line. It also gets reordered — by the same buyer, for the same line — for the high-temperature hydrocarbon service three flanges downstream, because the catalogue line is the same. Two years later, the high-temperature joint is blowing by and the next turnaround is extended by a week.

The four gasket families cover almost every service condition an industrial buyer will see, and matching the gasket to the service is the single highest-leverage decision on the joint.

Failure 4: Stud Length Off by One Thread Pitch

The stud is two threads short of the nut face after tightening. The fitter either tightens against the stud end — which is the wrong clamp load — or swaps in a longer stud from the bin, which is the wrong material or the wrong thread. Either way, the joint is not what the datasheet says it is, and there is no documentary trace of the substitution.

The standard rule, embedded in ASME PCC-1 and most EPC specifications, is that the stud must protrude a minimum of two thread pitches past the nut face after final tightening. That gives the next fitter visible proof the joint has been torqued, and it gives the joint enough thread engagement to hold the design clamp load through thermal cycling. The fix is to spec the stud length from the actual flange-and-nut combination on the datasheet — including washer thickness — not from a generic stud-bolt table.

A second, related failure is a hardness mismatch on NACE / sour service. B7 studs in H₂S service must be replaced by B7M (controlled hardness, HRC 22 max). A standard B7 stud will be rejected at the receiving dock, and the site will be stuck waiting for a replacement shipment while the system is drained. The spec line for stud material on any sour-service PO must read B7M per NACE MR0175, not just B7.

Failure 5: Single-Pass Torque to Final Value

The joint looks tight, the torque wrench reads 100% of target, and the crew moves on to the next flange. Two hours later, the gasket has relaxed, the clamp load has dropped, and the joint is weeping on first pressurization. This is the most common assembly-side failure, and it is also the easiest to fix.

The ASME PCC-1 four-pass star-pattern boltup is the standard procedure on every major refinery, FPSO, and chemical plant in the world, and it should be on every datasheet as a reference document. The four passes are not redundant — each one does a specific job.

• Pass 1 — 30% of target torque. Snug the gasket evenly around the bolt circle. Skipping this pass is the most common cause of a hydrotest failure.
• Pass 2 — 60% of target torque. Continue the star pattern. The gasket begins to flow into flange-face micro-imperfections.
• Pass 3 — 100% of target torque. Final seating pass. Calibrated torque wrench only — no impact guns on the final pass.
• Pass 4 — 100% re-pass after 4 hours (or first thermal cycle). Restores the design clamp load after gasket stress relaxation.

The Single-Source Advantage on a Bolted-Joint Package

Each of the five failures above is preventable with a tightening of the spec. The hard part is making sure all three sub-packages — the flange, the gasket, and the stud bolt and nut — arrive on site together, with documentation that lines up, and with no silent substitutions. A flange from one supplier, a gasket from a distributor, and a bag of studs from a fastener shop is the classic recipe for a joint that fails on first pressurization because nobody checked that all three submittals were compatible.

A single source for the entire joint package collapses that risk. The receiving inspection is faster, the MTC package is unified, and the engineering team has a single point of contact for any non-conformance. It also makes the inevitable scope changes during construction easier to manage — a flange class revision on the datasheet automatically pulls the matching gasket and the matching stud bolt through the same vendor.

How EZ Steel Industrial Builds a Bolted-Joint Package

Drawing on more than three decades in industrial pipe, tube, and component manufacturing, EZ Steel Industrial supplies the full bolted-joint package from a single source, with one MTC file and one project manager across the joint:

• Carbon-steel, stainless, alloy, and Cu-Ni pipe flanges to ASME B16.5, EN 1092-1, and GOST standards.
• Spiral-wound, ring-joint, kammprofile, and compressed-fibre gaskets to ASME B16.20 and B16.21, with full traceability on every ring.
• ASTM A193 / A320 stud bolts and ASTM A194 nuts, including B7, B7M, B8, B8M, B16, and L7 grades, plus matched hardened washers.
• Engineered kits for refinery, petrochemical, power, marine, and cryogenic service — including the matched industrial valves for the line.
• A track record that includes the South-to-North Water Diversion, the West-East Gas Pipeline, and overseas refinery, petrochemical, and shipbuilding contracts.