A leaking flange joint is rarely a manufacturing problem. The flange was machined correctly, the gasket met its specification, and the studs are the right grade. The leak happened because the joint was not assembled correctly, the gasket was wrong for the service, the bolt load was inadequate, or thermal cycling relieved the seating stress that was barely sufficient in the first place. These are design and assembly failures, and they are preventable.

Flange joint leaks in process plant carry significant consequences — lost product, environmental release, fire and explosion risk where the fluid is flammable, and unplanned shutdowns that cost multiples of what the joint repair costs. Understanding what makes a flange joint seal and what causes it to stop sealing is the foundation for preventing both the leak and its consequences.

How a Flanged Joint Works

A bolted flange joint seals by compressing a gasket between two flange faces with sufficient force that the gasket conforms to any surface irregularities and maintains contact pressure that exceeds the internal fluid pressure trying to escape. The bolt load must achieve two objectives simultaneously:

Both conditions must be satisfied simultaneously. A joint with sufficient bolt load to seat the gasket but insufficient to maintain it under pressure will leak when the system is pressurised. A joint with bolt load sufficient for pressure but insufficient to seat the gasket will leak from the first test.

Gasket Types and When to Use Them

Gasket selection is the most consequential design decision in a flanged joint. Using the wrong gasket type for the flange class, facing, or service conditions is a reliable route to a leaking joint regardless of how carefully the assembly is performed.

Compressed Non-Asbestos Fibre (CNAF)

CNAF sheet gaskets — cut from sheet material composed of fibres (glass, aramid, carbon) bound in a rubber or polymer matrix — are the general-purpose workhorse for raised face flanges in moderate service. They are inexpensive, easy to cut to size, and available in a wide range of compositions for different chemical services.

Limitations: CNAF is not suitable for high-pressure steam above approximately 40 bar, is subject to creep relaxation under sustained load (particularly at elevated temperature), and the quality between manufacturers varies significantly. The m and y values (see gasket parameters below) for CNAF vary considerably by grade — always use manufacturer data rather than generic tables.

Spiral Wound Gaskets (SWG)

Spiral wound gaskets consist of a V-shaped metal strip (typically 316L stainless or other corrosion-resistant alloy) wound in a spiral with a filler material (typically graphite or PTFE). They include an inner ring to prevent over-compression and — for standard flange applications — an outer centering ring that locates in the raised face bore.

SWGs offer better temperature and pressure performance than CNAF, lower creep relaxation, and higher reliability in cyclic service. They are the standard gasket for Class 300 and above in process pipework, and are widely specified at Class 150 where the service is demanding. The centering ring is critical — a spiral wound gasket without a centering ring on a raised face flange will migrate under compression and may not seal correctly.

Kammprofile (Grooved Metal) Gaskets

A solid metal core with concentric serrations machined into the face, covered with a soft facing material (typically graphite or PTFE). The serrations bite into the flange face under bolt load, creating a highly reliable metal-to-metal seal backed by the soft facing. Kammprofiles are used in high-temperature, high-pressure service, heat exchanger joints, and where fugitive emissions are regulated. They are more expensive than SWGs but more tolerant of flange face imperfections.

Ring Type Joint (RTJ) Gaskets

Solid metal rings — either oval or octagonal cross-section — that seat in machined grooves in the flange face. The ring is softer than the flange material and deforms into the groove under bolt load, creating a metal-to-metal seal. RTJ gaskets are used at Class 600 and above in high-pressure service, in hydrogen and sour service (NACE), and where the highest leak integrity is required. They require matching RTJ-faced flanges — RTJ gaskets cannot be used on raised face flanges.

PTFE and ePTFE

Full-face PTFE cut gaskets are used primarily with flat-face flanges (cast iron, equipment nozzles) and in chemical service where stainless or graphite-containing gaskets are not compatible. Expanded PTFE (ePTFE) tape, wrapped in the gasket groove, is used for low-pressure service and provides excellent chemical resistance. Neither is suitable for elevated temperatures above approximately 200°C or high bolt loads that will cause excessive cold flow.

Gasket Parameters — m and y Values

The ASME pressure vessel and piping codes characterise gasket sealing performance using two parameters, used in the bolt load calculation:

Gasket typem (factor)y (MPa min)Notes
PTFE (full face)0.5 – 1.01.4 – 2.8Low seating requirement, verify cold flow
CNAF (rubber sheet)1.0 – 2.02.8 – 11.0Wide range — use manufacturer data
CNAF (fibre-reinforced)2.0 – 3.011.0 – 25.0Grade-dependent — confirm with supplier
Spiral wound (graphite fill)3.031.0Per ASME B16.20 standard values
Spiral wound (PTFE fill)2.5 – 3.020.0 – 31.0Lower m than graphite in most grades
Kammprofile (graphite faced)3.0 – 4.040.0 – 55.0Manufacturer data essential
RTJ oval/octagonal5.5 – 6.5124 – 179High seating — proportionally high bolt load required
Critical note on m and y values: The ASME m and y values in Appendix 2 of ASME VIII are guidance values only, and several major gasket manufacturers publish data showing significant deviation from the ASME table values for their specific products. For any joint where reliability is critical, obtain m and y data from the gasket manufacturer and use those values in the bolt load calculation rather than the generic code tables.

Bolt Load Calculation

The required bolt load for a flanged joint is calculated from two conditions, and the greater controls the design:

Gasket Seating Condition (Wm2)

Wm2 = π × b × G × y

Where b is the effective gasket seating width (mm), G is the mean gasket diameter (mm), and y is the minimum seating stress (MPa). This is the bolt load required to seat the gasket before pressurisation.

Operating Condition (Wm1)

Wm1 = H + Hp

Where H is the hydrostatic end force (the pressure force acting to separate the flanges = π/4 × G² × P) and Hp is the compression load required to maintain the gasket seal under pressure = 2b × π × G × m × P.

The available bolt load from the specified studs must exceed the greater of Wm1 and Wm2. Available bolt load = number of bolts × root area of bolt × allowable bolt stress at temperature.

If the available bolt load is inadequate, the options are: increase bolt size, increase the number of bolts (not possible with standard flanges without modifying the flange), change to a higher allowable stress bolt material (B7 rather than B8), or change the gasket to one with lower m and y values.

Effective Gasket Width

The effective seating width b is not simply the full gasket face width. The ASME code defines an effective width based on the contact geometry:

This reflects the fact that wider gaskets do not seat uniformly — the inner and outer edges carry more load than the centre, and the effective sealing area is less than the total gasket area. Specifying unnecessarily wide gaskets does not improve the seal — it increases the required bolt load while reducing the effective seating stress across the face.

Thermal Cycling and Bolt Load Relaxation

A joint that seals satisfactorily at ambient temperature during hydraulic test may leak on first heat-up. This is one of the most common — and most avoidable — flange joint failures.

The mechanism is straightforward: the gasket material (particularly CNAF and PTFE) experiences creep relaxation under sustained compressive load. When the system heats up, the gasket and bolt materials expand thermally. If the thermal expansion coefficients of the gasket, flange, and bolt differ significantly (and they almost always do), the bolt load changes. In most configurations, the gasket creeps under the hot conditions and the bolt elongation due to thermal expansion partially relieves the bolt tension — the net result is that the joint loses bolt load.

Practical implications:

Assembly — Where Most Leaks Begin

The majority of flange joint leaks trace to assembly errors rather than design or material failures. The most critical assembly requirements:

Flange Face Condition

Raised face flanges must be inspected before assembly. Surface finish should be in the range Ra 3.2–6.3 μm for standard gaskets — smooth enough for gasket conformance but with sufficient texture for the gasket to grip. Radial scratches are significantly more damaging than circumferential marks because they create a potential leak path. Deep scratches, weld spatter, corrosion pitting, or impact damage on the seating face must be assessed — a damaged face may need refinishing or replacement before a reliable seal can be achieved.

Gasket Handling and Positioning

Gaskets must be centred on the flange face and positioned correctly before bolt-up. A spiral wound gasket installed without its centering ring, or a cut gasket that is off-centre covering the bolt holes, will not seal correctly regardless of bolt load. Gaskets must not be reused — even if visually undamaged, a used gasket has permanently deformed and will not provide the same seating performance as a new one.

Bolt Tightening Sequence and Method

Bolts must be tightened in a cross-bolt pattern — opposing bolts alternately — rather than sequentially around the circle. Sequential tightening lifts the opposite side of the joint and causes uneven gasket loading that never fully equalises. The standard procedure per ASME PCC-1 is:

  1. Hand-tighten all nuts snug
  2. Tighten to 30% of target torque in cross pattern
  3. Tighten to 70% of target torque in cross pattern
  4. Tighten to 100% of target torque in cross pattern
  5. Final pass clockwise around the bolt circle to confirm no further rotation

Torque vs Tension

Bolt load is controlled by torque in the vast majority of installations. The relationship between applied torque (T) and achieved bolt tension (F) is:

T = K × d × F

Where K is the nut factor (typically 0.15–0.20 for lightly oiled or lubricant-coated threads, up to 0.25 for dry threads), d is the nominal bolt diameter, and F is the bolt tension. The scatter in achieved bolt tension from torque tightening is typically ±25–30% even under controlled conditions — primarily due to variation in the friction coefficient K.

This scatter means that if the target bolt load is calculated assuming K=0.20, some bolts will actually be tightened to 75% of that load and some to 125%. The design calculation must account for this scatter — the minimum bolt load (accounting for under-tightening) must still be sufficient to seat the gasket, and the maximum bolt load (accounting for over-tightening) must not crush the gasket or exceed bolt yield.

For high-integrity joints — large bore, high pressure, hazardous fluids — hydraulic bolt tensioners are used instead of torque tools. Tensioners apply a direct axial load to the stud (bypassing friction entirely) and achieve bolt load scatter of ±5–10%, significantly improving joint reliability.

EN 1591-1 — The European Calculation Method

EN 1591-1 provides a more rigorous bolt load calculation method than the ASME Appendix 2 approach, accounting for flange rotation, bolt relaxation, and the interaction between flange stiffness and gasket compression in a more complete way. It requires more input data — particularly gasket stress-compression characteristics from EN 13555 — but produces a more accurate assessment of joint behaviour and is particularly valuable for non-standard geometries.

Where a joint calculation is required for conformity under EN 13480 (European piping standard) or EN 13445 (pressure vessels), EN 1591-1 is the appropriate calculation framework. For ASME B31.3 pipework, ASME Appendix 2 is the standard approach.

Common Failure Modes — A Diagnostic Reference

Failure modeSymptomsCausePrevention
Under-tighteningLeaks from first pressurisationInsufficient torque, wrong K factor, no cross-bolt sequenceTorque specification, calibrated tools, correct procedure
Gasket crushGasket expelled or deformed, leaks on retighteningOver-tightening, no inner ring on SWGMaximum torque limit, correct gasket specification
Thermal relaxationLeaks on heat-up or after first cycleGasket creep at temperature, differential thermal expansionSpecify graphite-filled SWG, re-tighten after first heat-up
Wrong gasket face typeRF gasket on FF flange — cracked flange, persistent leakSpecification error or site substitutionVerify gasket type against flange face type before assembly
Damaged flange faceLocalised leakage path, does not respond to retighteningImpact, corrosion, weld spatter, improper previous disassemblyFace inspection before assembly, do not reuse damaged gaskets
MisalignmentUneven bolt loads, gasket distortion, persistent single-side leakPipework not supported before joint make-upAlign pipework before bolting, support independently of joint
Gasket reuseImmediate or early leak on reassembly after maintenancePermanent deformation of used gasketAlways fit a new gasket after joint disassembly

ASME PCC-1 — The Joint Assembly Standard

ASME PCC-1 (Guidelines for Pressure Boundary Bolted Flange Joint Assembly) is the primary reference for flange joint assembly procedures. It covers: pre-assembly inspection, gasket handling, bolt lubrication, tightening sequence, torque tool calibration, hot torquing, and documentation requirements.

For safety-critical joints — hydrogen service, toxic fluids, high-pressure steam — many operators require documented joint assembly in accordance with PCC-1, with the joint assembler signing off that the procedure was followed. This creates an audit trail and significantly improves the probability of a leak-free joint. The incremental cost of applying PCC-1 to critical joints is small relative to the cost of a leak.

Summary

A flanged joint leaks when the residual gasket contact stress falls below the level required to maintain the seal — either because the initial bolt load was insufficient, the gasket relaxed, or the bolt load was relieved by thermal cycling or fluid pressure. All three failure mechanisms are predictable and can be designed and assembled against.

The design responsibilities are: select the correct gasket type for the flange facing, service conditions and pressure-temperature duty; calculate the required bolt load from the gasket m and y values using ASME Appendix 2 or EN 1591-1; confirm the specified studs can deliver that bolt load; and specify the torque values and tightening procedure on the joint documentation. The assembly responsibilities are: use those torque values, in the correct sequence, with calibrated tools, on a correctly prepared face, with a new gasket centred correctly.

Most persistent flange joint leaks on operating plant can be resolved without replacing equipment. Inspect the face condition, replace the gasket with an appropriate new gasket, retorque to specification in the correct sequence, and if the service is cyclic or elevated temperature, plan a re-torque after the first heat cycle.

Forgepoint provides process pipework design including flange specification, gasket selection, and bolt load calculations. If you need engineering support on a pressure system, get in touch.

Discuss Your Project — 07549 032776