{"id":5584,"date":"2025-12-01T10:11:30","date_gmt":"2025-12-01T02:11:30","guid":{"rendered":"https:\/\/sunhyings.com\/?p=5584"},"modified":"2026-01-30T15:21:11","modified_gmt":"2026-01-30T07:21:11","slug":"hydrogen-flange-leakage-solutions","status":"publish","type":"post","link":"https:\/\/sunhyings.com\/ar\/blog\/hydrogen-flange-leakage-solutions\/","title":{"rendered":"\u0641\u0644\u0646\u062c\u0627\u062a \u0627\u0644\u0647\u064a\u062f\u0631\u0648\u062c\u064a\u0646 \u0639\u0627\u0644\u064a\u0629 \u0627\u0644\u0636\u063a\u0637: \u062d\u0644\u0648\u0644 \u0644\u0645\u0646\u0639 \u0627\u0644\u062a\u0633\u0631\u0628 \u062a\u0645\u0627\u0645\u064b\u0627"},"content":{"rendered":"\n
\"Solutions
Hydrogen flange joints fail most often from a combined issue: insufficient bolt load, unstable gasket stress, and material\/finish choices that do not match H2<\/sub> permeation and hydrogen-assisted damage mechanisms.<\/figcaption><\/figure>\n\n\n\n

Stopping flange leakage<\/a> in high-pressure hydrogen service requires more than \u201ctightening bolts.\u201d You need a joint design and assembly method that controls bolt load (not just torque), uses hydrogen-appropriate materials, and applies a gasket system with proven low leakage under cycling.<\/strong><\/p>\n\n\n\n

Hydrogen is challenging to seal because it is a very small, fast-diffusing molecule in many non-metallic materials. A practical way to explain this is kinetic diameter: H2<\/sub> is typically listed around 2.89 \u00c5<\/strong> versus methane around 3.80 \u00c5<\/strong>, so the sealing system has less margin against micro-paths and permeation through some polymers and composites. If your service is in the \u201cvehicle-fueling class\u201d (often 35 MPa \/ 70 MPa<\/strong>, i.e., 350 bar \/ 700 bar), even a micro-leak can become a safety and availability event. For code compliance, the baseline reference in many projects is ASME B31.12 (Hydrogen Piping and Pipelines)<\/a>, with bolted-flange assembly guidance commonly aligned to ASME PCC-1<\/a>. When you build your approach around bolt load, gasket stress, and verified surfaces, leakage control becomes repeatable\u2014not luck.<\/p>\n\n\n\n

Root Causes of Flange Leakage in Hydrogen Service<\/h2>\n\n\n\n

Flange leakage in high-pressure hydrogen service happens because multiple mechanisms stack up: bolt-load loss (relaxation), uneven gasket stress, surface finish defects that become leak paths, and hydrogen-assisted material damage in susceptible alloys\/fasteners.<\/strong><\/p>\n\n\n\n

Unlike natural gas, hydrogen can participate in hydrogen-assisted cracking mechanisms in certain steels and high-strength fasteners, especially when strength\/hardness is high or when plating\/process control is poor. Separately, at elevated temperature, some carbon steels can suffer High Temperature Hydrogen Attack (HTHA)<\/em> (typically managed using API RP 941 (Nelson Curves)<\/a>). In many ambient-temperature high-pressure H2<\/sub> systems (compressors, storage, dispensing), the most common day-to-day leak drivers are still mechanical: bolt load scatter, flange rotation, gasket seating\/creep, and surface damage. Understanding these root causes is central to applying ASME B31.12<\/strong> correctly and defensibly.<\/p>\n\n\n\n

Cause of Leakage<\/th>Description<\/th><\/tr>
Inadequate Bolt Preload<\/td>Low initial bolt load cannot maintain gasket stress against internal pressure. In hydrogen, \u201cborderline\u201d preload tends to show up as micro-leaks first, then grows with cycles.<\/td><\/tr>
Improper Bolt Tightening Sequence<\/td>Uneven tightening causes local gasket over-compression and under-compression. Under-compressed sectors become leak paths during pressurization and thermal cycling.<\/td><\/tr>
Poor Contact Surface Quality<\/td>Radial scratches, dents, or an out-of-spec finish create direct leak channels. For many ASME B16.5 raised faces with spiral-wound gaskets, typical stock finish is in the 125\u2013250 \u00b5in Ra<\/strong> range (roughly 3\u20136 \u00b5m Ra) when properly machined and undamaged.<\/td><\/tr>
Hydrogen-Assisted Damage (HE Risk in Susceptible Parts)<\/td>High-strength\/hard fasteners and certain steels are more susceptible. Hydrogen effects increase with strength\/hardness and with stress concentration at threads or under nut bearing surfaces.<\/td><\/tr>
Gasket Construction Not Matched to H2<\/sub> Service<\/td>Some non-metallic\/composite systems can allow higher permeation or interface leakage under cycling. For very high pressure, metal-to-metal style sealing (e.g., RTJ) is often selected; otherwise use gasket constructions with strong recovery and proven leakage performance under cycling.<\/td><\/tr>
Bolt Surface Condition & Friction Scatter<\/td>Dirty threads, galling, mixed lubrication, or re-used nuts change friction. Torque then becomes a poor proxy for bolt load, creating clamp-load scatter and leak risk.<\/td><\/tr>
Misalignment \/ Flange Rotation<\/td>Angular misalignment and piping loads bend the joint, unloading the gasket on one side. This is a frequent \u201cmystery leak\u201d driver after maintenance or pipe modification.<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n

Bolt Preload & Relaxation<\/h3>\n\n\n\n

Incorrect bolt preload is a leading cause of flange leakage.<\/strong>
In high-pressure service, bolts commonly lose effective clamp load early because the gasket seats (embedment) and relaxes. That bolt-load loss can happen within hours to the first day after assembly, and then again after the first thermal cycle. ASME PCC-1<\/a> treats bolted flange joints as a system: flange stiffness, gasket type, lubrication condition, and tightening method all matter.<\/p>\n\n\n\n

Engineering control that actually works:<\/strong> treat torque as an indirect method. For critical hydrogen joints, reduce scatter by standardizing lubricant, nut\/washer condition, and tool calibration, and consider bolt-load verification (bolt elongation, ultrasonic measurement) when feasible. If you must use torque, document the assumed friction factor and keep it consistent across the joint.<\/p>\n\n\n\n

Field example (representative):<\/strong> A Class 900 raised-face joint passed a soap test at low pressure but developed a micro-leak after first pressurization and cooldown. The root cause was mixed lubrication: half the studs were lightly oiled, half were dry from storage. Torque values were identical, but clamp load was not. The corrective action was a controlled clean\/lube procedure (same lubricant, same washer type), a multi-pass tightening sequence, and a torque check pass after the first temperature stabilization.<\/p>\n\n\n\n

Improper Tightening Sequence<\/h3>\n\n\n\n
\"\"<\/figure>\n\n\n\n

Improper tightening sequence leads to uneven gasket compression and flange leakage.<\/strong>
You should tighten bolts in a Legacy Star Pattern (Cross Pattern) to distribute gasket stress. For most joints, a multi-pass approach is more stable than a single \u201cfinal torque\u201d pass. A common controlled method (aligned with industry practice reflected in ASME PCC-1<\/a>) is: ~30% \u2192 ~60% \u2192 100%<\/strong> of target torque in cross pattern, followed by a circular check pass<\/strong> at 100% until nuts stop turning.<\/p>\n\n\n\n

Practical note for hydrogen:<\/strong> if a joint leaks, \u201crandom retightening\u201d often makes it worse by creating localized gasket crush. Retightening should be controlled, patterned, and documented\u2014or the joint should be opened and corrected (alignment, surface, gasket damage) rather than forced.<\/p>\n\n\n\n

Surface Finish Requirements<\/h3>\n\n\n\n
\"\"<\/figure>\n\n\n\n

Poor surface finish or damage increases the risk of fugitive emissions.<\/strong>
For many ASME B16.5 raised-face applications using spiral-wound gaskets, the commonly referenced stock finish range is 125\u2013250 \u00b5in Ra<\/strong> (about 3\u20136 \u00b5m Ra). What matters most in hydrogen is not \u201cpolish,\u201d but no radial scratches<\/strong>, no dents, and a consistent machining pattern. A concentric serrated finish is often preferred when you are trying to avoid \u201cstraight-through\u201d leak channels. If your face has radial scoring, do not \u201ctorque your way out\u201d of it\u2014re-machine or replace the flange.<\/p>\n\n\n\n

Field example (representative):<\/strong> A joint repeatedly leaked at the same clock position despite gasket replacement. The actual driver was a shallow radial tool mark crossing the raised face. The fix was a controlled re-facing to the correct roughness range and a final inspection using lighting at a low angle to reveal radial features. The repeat leak stopped without increasing torque.<\/p>\n\n\n\n

Corrosion & Hydrogen Attack<\/h3>\n\n\n\n

Corrosion and hydrogen-related damage can weaken sealing surfaces and create new leak paths.<\/strong>
Two different issues are often mixed up:<\/p>\n\n\n\n