Boiler pressure parts — fabrication, repair and replacement

Pressure parts are the boiler's pressure boundary: everything that contains water or steam above atmospheric pressure. Over a boiler's 25–40 year life, most of that boundary is replaced at least once — erosion from abrasive bagasse and rice-husk ash, fireside corrosion, and creep in the hottest sections see to it. Our fabrication scope, part of the wider boiler products line, covers:

• Headers — superheater, economizer and water-wall headers, drilled, stub-welded and end-capped, in carbon and alloy grades up to SA 335 P22 pipe and equivalent forgings.
• Drum repair scope — new internals (cyclone separators, chevron driers, feed distribution pipes), nozzle build-up and weld repairs to the shell under the owner's inspection authority. We do not fabricate new drums; drum-shell work is repair and internals only.
• Tube bundles — superheater and economizer bundles, bent, welded to headers, and shipped as liftable modules.
• Membrane panels — water-wall panels, tube-fin-tube welded, delivered flat or pre-formed with openings for burners and soot blowers.
• Coils — economizer and heat-recovery coils, plain or finned tube.

Each delivery is engineered against the boiler it enters: tube pitch and bend radii from the original arrangement or our own dimensional survey, materials verified rather than assumed, and welds prepared so site tie-in joints land in accessible positions.

Fabrication runs at our factory and R&D centre in Samut Sakhon, Thailand:

• Tube and pipe bending — mandrel and induction bending with ovality and wall-thinning checks against code limits (wall thinning at the extrados held within the allowance assumed in the thickness calculation).
• Welding to ASME Section IX — qualified procedures (WPS/PQR) and qualified welders for each material group and process; GTAW roots on all tube butt welds; alloy consumables stored and issued under heated-quiver control.
• Post-weld heat treatment — furnace and local-resistance PWHT with recorded time–temperature curves; typical soak 705–745 °C for T22/P22 headers, with T91 treated in its tighter code-mandated band.
• Hydrostatic testing — every pressure part tested at 1.5 × design pressure per the code of construction, held and inspected before draining and drying for shipment.

Traceability is 100 %: every tube, plate and forging enters the works with a mill test certificate (MTC), is checked against the specification — chemistry, mechanicals, heat treatment condition, and the certificate's endorsement (EN 10204 type 3.1, or 3.2 with an independent inspector's counter-signature where the purchaser's specification requires it) — and its heat number is transferred to every cut piece and recorded on the as-built weld map. Positive material identification by XRF backs up the paperwork on alloy grades, because a single T11 stub in a T22 header is invisible to the eye and fails in service. The final dossier — MTCs, WPS/PQR, welder qualifications, NDT reports, PWHT charts, hydrotest certificate, as-built drawings — is what the owner's inspector and insurer actually buy along with the steel.

Should failed superheater sections be replaced with the same material?

Only if the original grade was right for the measured metal temperature. Remaining-life assessments repeatedly find the same pattern: a section specified to the steam temperature, running 30–60 °C hotter in the metal, oxidising and creeping years ahead of schedule. Re-engineering that section — T11 to T22, or T22 to T91 — removes the mechanism; replacement in kind merely restarts the clock on it.

Our engineering step between survey and fabrication therefore asks three questions of every worn section. What was the actual metal temperature — from oxide-scale thickness measurement and hardness survey, not the design sheet? What was the failure mechanism — long-term overheating shows thick-lipped creep fracture and heavy scale; erosion shows polished, thinned flats facing the gas lane; under-deposit corrosion shows pitting at the water side? And does the boiler's current operation match its design — many bagasse units now run harder, with higher final steam temperature, than their 1990s design basis assumed? Where the answers say the original specification was wrong, we propose the upgrade with the arithmetic attached; where they say the wear is simply end-of-life, replacement in kind is the cheaper and correct answer. Erosion-driven cases often justify shields and flow baffles rather than a grade change. The same logic extends to geometry: gas-side lane widths and fin pitch can be corrected at replacement to cut the erosion velocity that killed the original bundle.