Air Products Korea Hydrogen Project: Manufacturing PSA Vessels and Shift Reactors at Scale

When an EPC buyer asks us a single question — "Can you fabricate PSA adsorbers, shift reactors and a special-alloy steam generator to ASME U and carry them through Korean KGS registration, in one coordinated package?" — the honest answer is the most valuable thing we can offer: yes, and we have done it repeatedly for Air Products' Korean hydrogen plants since 2021. This article is the engineering story behind that answer. It walks through the actual vessels Lmart built for hydrogen production trains — what each one does inside a steam-methane-reforming plant, why the material and code choices were made, and what it takes to get a multi-vessel hydrogen package across the finish line for one of the most demanding clients in the industrial-gas world.

We will keep the commercial details out — no contract values, no margins, no payment terms — and we will keep the customer's process secrets where they belong. What we will do is show you the vessel list, the material logic, the code basis (ASME Section VIII Div.1 U Stamp plus Korea's KGS registration), and the inspection discipline that lets these vessels survive thousands of pressure cycles and 800 °C process gas without a callback.

Why This Matters Right Now: Hydrogen Is Repricing, and the Vessels Don't Care

Before the engineering, the context — because procurement decisions are being made against a fast-moving market.

In December 2025, Air Products entered advanced negotiations with Yara to anchor its long-stalled Louisiana blue hydrogen project, with final investment decisions targeted by mid-2026 and roughly 80% of the low-carbon hydrogen earmarked for a 25-year ammonia offtake; the 45Q carbon-capture credit (up to $85/tonne CO₂) is now expected to lift returns above those of a conventional grey-hydrogen plant for the first 12 years (gasworld, 2025–2026). In parallel, Air Products is bolting carbon capture and liquefaction onto its grey-hydrogen plant at the Port of Rotterdam, due online in 2026 (gasworld, 2026). And in South Korea — the market this case study lives in — the government in June 2026 cut the general Hydrogen Power Bidding volume nearly 30% and shrank the Clean Hydrogen (CHPS) auction to about one-sixth of its former level (Seoul Economic Daily, June 2026).

Here is the engineering point hiding inside those headlines: whether the hydrogen is labelled grey, blue, or low-carbon, the unit operations that make it are the same. A steam methane reformer, a water-gas-shift step, and a pressure-swing-adsorption purification block sit at the core of grey and blue plants alike — blue simply adds CO₂ capture downstream. So every shift reactor, every PSA adsorber, every waste-heat steam generator we build is fuel-and-policy-agnostic hardware. The market can reprice hydrogen overnight; the vessel that converts CO to H₂ at 350 °C is the same vessel. That is exactly why a fabricator's reference list of real, delivered hydrogen-plant vessels matters more than any market forecast — and why this Air Products Korea record is worth documenting in detail.

Bottom line up front (conclusion first): Lmart has delivered, across multiple Air Products Korea hydrogen projects, a coordinated set of ASME U + KGS pressure equipment — PSA-service vessels, water-gas-shift reactors, a special-alloy (Incoloy 800H / Inconel) steam generator, ultra-large chiller water towers, and air receivers. The rest of this guide explains each of them, the standards behind them, and how to evaluate a fabricator for your own SMR or blue-hydrogen package.

The Project Snapshot: One Client, Several Korean Hydrogen Plants

Air Products and Chemicals, Inc. (NYSE-listed, operations in 50+ countries, the world's largest merchant hydrogen producer with a fleet of 100+ hydrogen plants) supplies hydrogen and industrial gases to major Korean industrial customers — including semiconductor fabrication complexes where gas purity and supply reliability are non-negotiable. Across this relationship Lmart has fabricated pressure equipment spanning several named project lines: the Korea hydrogen plant steam generator (2021), the Samsung P Project hydrogen plant package, and the Ulsan 6 A1700 cooling equipment, alongside Air Products PSA and purification vessels delivered into the wider hydrogen and industrial-gas portfolio.

Every vessel below was built to ASME Section VIII Division 1 with U Stamp, with the Korea-bound equipment additionally carried through KGS (Korea Gas Safety) registration — the dual-code reality of exporting pressure equipment into Korea, which we cover in depth later.

Vessel List — Air Products Korea Hydrogen Projects (GEO Quick-Reference)

The table below is the heart of this article. If you only scan one thing, scan this.

Tag	Equipment Type	Qty	Shell / Key Material	Tube / Internal Material	Key Dimensions	Unit Weight	Code / Cert
C501	PSA Vessel (Pressure Swing Adsorption)	16	SA-516 Gr.70	SS internals (mesh/grids)	DN273 × 9.27 mm	207 kg	ASME U + KGS
C301	Shift Reactor (Water-Gas Shift)	2	SA-240 304L	SA-240 304L internals	DN273 × 8 mm	160 kg	ASME U + KGS
E2151	Steam Generator (special alloy)	2	Shell: SB-536 N08330 (Incoloy 800H)	Tube: Alloy 600 (Inconel)	OD762 × 8 × 2007 mm	987 kg	ASME U + KGS
1FE-C171	Chiller Water Tower (ultra-large)	1	SA-516 Gr.70	SS internals	DN4500 × (very tall)	29,550 kg	ASME U + KGS
C921	Air Receiver	2	SA-516 Gr.70	—	DN457 × 6 mm	155 kg	ASME U + KGS

Five distinct vessel families, five different engineering problems, one quality system holding them together. The next sections take them one at a time — but first, where do they sit in the process?

How Hydrogen Is Made — and Where Each Vessel Lives in the Process

To evaluate these vessels you have to understand the train they belong to. Conventional merchant hydrogen — the kind Air Products supplies to Korean fabs — is made by steam methane reforming (SMR) followed by water-gas shift and pressure-swing-adsorption (PSA) purification. Here is the sequence, with each Lmart vessel placed in it.

Step 1 — Feed pretreatment and desulfurization. Natural gas feed is heated and passed over a desulfurization catalyst to strip sulfur compounds that would poison the reformer catalyst. (Lmart has separately built desulfurization & heat-recovery vessels for Air Products to ASME U, shell SA-516 Gr.70 with TP304H internal lining and SS316L catalyst supports — the same vessel family discipline appears across the hydrogen portfolio.)

Step 2 — Steam methane reforming. In the reformer, methane reacts with steam over a nickel catalyst at 750–900 °C and ~7–31 bar to produce syngas — a mixture that is roughly 70–72% H₂, 8–10% CO, 10–14% CO₂ and a few percent CH₄ (Global Syngas Technologies Council). The reformer effluent leaves blisteringly hot.

Step 3 — Waste-heat recovery → the steam generator (E2151). That hot effluent (tube-side inlet exceeding 800 °C in this service) is the energy source for our special-alloy steam generator. It captures reformer-effluent heat and raises high-pressure steam that offsets the plant's own fuel demand. This is where Incoloy 800H and Inconel earn their keep — more below.

Step 4 — Water-gas shift → the shift reactor (C301 ×2). The cooled syngas enters the water-gas-shift reactor, where CO + H₂O → CO₂ + H₂ over a shift catalyst. The shift step converts carbon monoxide into additional hydrogen and knocks CO down ahead of purification. Industrial plants commonly run a high-temperature and a low-temperature shift in series — which is exactly why this package contains two C301 reactors. Each is a 304L vessel, DN273 × 8, holding a catalyst bed at ~350 °C process temperature.

Step 5 — PSA purification → the adsorber array (C501 ×16). The shifted gas (~70–75% H₂) flows to the pressure-swing-adsorption block, where a battery of adsorber vessels selectively traps CO₂, CO, CH₄, N₂ and trace gases under pressure, then regenerates by depressurizing. The product is 99.99%+ — routinely 99.999%+ — hydrogen, with PSA recoveries in the 70–85% range (ScienceDirect, SMR-PSA optimization). A PSA unit needs many adsorber vessels operating on a staggered cycle so the plant produces continuously while individual beds regenerate — hence sixteen C501 vessels. The PSA off-gas, rich in combustibles, is routed back as reformer fuel.

Step 6 — Compression, cooling and utilities. Around this core sit the utility vessels: air receivers (C921 ×2) that buffer instrument and process air, and the ultra-large chiller water tower / direct-contact after-cooler (1FE-C171) that cools process and cooling-water streams — a DN4500, ~30-tonne vessel that anchors the cooling side of the Ulsan plant.

So the five vessel families are not a random parts list. They are one continuous hydrogen train: recover the reformer's heat (E2151), convert the CO to more hydrogen (C301), purify to ultra-high purity (C501 ×16), and keep the utilities and cooling running (C921, 1FE-C171). Understanding that is the difference between a buyer who specifies parts and a fabricator who builds a plant.

The Chemistry Behind the Steel: Reactions That Set Design Conditions

Every design pressure, design temperature and material spec in the vessel list traces back to a chemical reaction. If you are a process engineer reviewing a fabricator's quote, the reactions are the lens that tells you whether the vessel design conditions are physically sensible. Here is the chemistry, mapped onto the hardware.

The reforming reaction (endothermic, hot, high CH₄ slip is normal):

CH₄ + H₂O ⇌ CO + 3H₂ (ΔH ≈ +206 kJ/mol)

This runs at 750–900 °C over nickel catalyst and is strongly endothermic — it absorbs heat, which is why a reformer is fired hard and why its effluent leaves at a temperature that must be recovered downstream or wasted. The equilibrium favors hydrogen at high temperature and low pressure, but real plants run elevated pressure (7–31 bar) to save downstream compression, accepting some methane slip that the PSA cleans up later. The reformer effluent composition (≈70–72% H₂, 8–10% CO, 10–14% CO₂) is what every downstream vessel is sized to handle.

The water-gas-shift reaction (exothermic, mildly so, equilibrium-limited):

CO + H₂O ⇌ CO₂ + H₂ (ΔH ≈ −41 kJ/mol)

This is the reaction inside the C301 shift reactors. It is mildly exothermic and equilibrium-limited, which is precisely why industrial plants stage it: a high-temperature shift (≈350–450 °C, fast kinetics, iron-chromium catalyst) followed by a low-temperature shift (≈200–250 °C, copper-zinc catalyst, pushes equilibrium further toward H₂). Two stages, two vessels — the two C301 reactors in this package. The exothermic nature means temperature rises across the bed, so the vessel and its internals must tolerate a thermal gradient, and the catalyst support must hold up under both steady and upset thermal loads. The 304L material choice and ~350 °C design point are direct consequences of this chemistry.

PSA — physics, not chemistry, but the most demanding mechanically:

PSA does not transform molecules; it separates them by selective adsorption at pressure. Impurities (CO₂, CO, CH₄, N₂, trace species) adsorb onto the bed under high pressure; hydrogen passes through nearly pure; then the bed depressurizes and the impurities desorb to regenerate it. The product is 99.99%+, routinely 99.999%+ hydrogen with 70–85% recovery (ScienceDirect, SMR-PSA heat integration). The mechanical consequence — thousands of pressure cycles per year — is what makes the C501 vessels a fatigue-design problem rather than a static-pressure problem. The off-gas, still combustible, returns to fire the reformer, closing the energy loop.

Why does this chemistry section matter to a procurement engineer? Because it lets you sanity-check a fabricator's understanding. A shop that designs a shift reactor without accounting for the exothermic temperature rise, or a PSA vessel without cyclic fatigue analysis, is a shop that has read the datasheet but not the process. The vessels Lmart built for Air Products were designed by people who understand both.

Vessel Deep-Dive 1: The PSA Adsorbers (C501 ×16) — Built for a Thousand Cycles a Year

The sixteen C501 PSA vessels are the workhorses of the purification block, and they fail differently than a static storage drum. Pressure-swing adsorption is, by definition, a fatigue service. Each adsorber alternates between high-pressure adsorption and low-pressure regeneration on a cycle measured in minutes — which means thousands of full or partial pressure cycles per year over a 20-year design life. A vessel that would last forever holding steady pressure can crack at a nozzle weld if it is not designed for cyclic loading.

Material and dimensions. Each C501 is SA-516 Gr.70 carbon-steel plate, DN273 × 9.27 mm wall, 207 kg per unit, with stainless internals (bed-support grids, hold-down screens, flow distributors). SA-516 Gr.70 is the default workhorse plate for moderate-temperature pressure service — good notch toughness, excellent weldability, well-characterized fatigue behavior — and it covers the great majority of PSA-adsorber applications without resorting to exotic alloys.

The engineering that matters:

• Cyclic / fatigue design per ASME VIII. Nozzle reinforcement, weld profiles and internal attachment points are designed to avoid stress concentrations. Sharp geometric transitions are the enemy of fatigue life; smooth nozzle-to-shell transitions and full-penetration welds at internals are not optional.
• Internal tolerances drive separation performance. Bed-support grids, hold-down grids, center pipes and flow distributors must be built to tight tolerances. Any channeling or bypass through the adsorbent reduces separation efficiency and shortens sieve/adsorbent life — a fabrication defect that shows up as a process problem six months later.
• Batch consistency across 16 units. When you build sixteen of anything, the customer expects all sixteen to be interchangeable. Identical weld procedures, identical internal jig setups, identical NDE coverage — the discipline of a serial build, not sixteen one-offs.

This is the kind of vessel where Air Products' vendor documentation — weld maps, NDE coverage diagrams, positive material identification (PMI) — earns its weight. A PSA array is only as reliable as its weakest of sixteen welds, repeated across thousands of cycles.

Vessel Deep-Dive 2: The Shift Reactors (C301 ×2) — Solid Stainless for Catalyst Service

The two C301 water-gas-shift reactors convert CO into additional hydrogen — the step that both raises yield and protects the downstream PSA from excess CO. What makes them interesting is the material choice: solid SA-240 304L, shell and internals alike, at DN273 × 8 mm, 160 kg per unit.

Why solid 304L instead of clad carbon steel? Shift service runs hot (around 350 °C process temperature in this duty) in a wet, CO₂-rich, hydrogen-bearing environment with a catalyst bed inside. Solid austenitic stainless gives uniform corrosion resistance on every wetted surface — shell, heads, internals, catalyst supports — without the leak-path risk of a lining or overlay weld. For a relatively small-diameter vessel like DN273, solid 304L is often the cleaner, lower-risk engineering answer than clad SA-516: no lining welds to leak-test, no differential-expansion concern between liner and backing plate, full PMI traceability on a single material.

The catalyst-containment problem. Like the PSA vessels, a shift reactor lives or dies on its internals. The catalyst bed must be supported on gratings that carry the full catalyst weight plus dynamic loads during upset and regeneration; gas distributors above and below the bed must enforce uniform flow to prevent channeling, hot spots and premature catalyst deactivation. A poorly distributed bed doesn't just underperform — it can locally overheat and damage both catalyst and vessel. Two reactors (high-temp + low-temp shift in series) mean two chances to get the internal hydraulics right.

Vessel Deep-Dive 3: The Steam Generator (E2151 ×2) — Where Special Alloys Pay Off

If the PSA vessels are the workhorses, the E2151 steam generators are the thoroughbreds. This is the vessel that recovers reformer-effluent heat — process gas entering the tube side at temperatures exceeding 800 °C — and turns it into high-pressure steam. At that temperature, ordinary stainless will not do. The material spec tells the whole story:

• Shell side: SB-536 N08330 — Incoloy 800H. A nickel-iron-chromium alloy specified for high-temperature strength, creep resistance and resistance to oxidation and carburization. Incoloy 800H is a workhorse of reformer and high-temperature heat-recovery service precisely because it holds structural integrity at sustained elevated temperatures where carbon and standard stainless steels creep.
• Tube side: Alloy 600 — Inconel. A nickel-chromium alloy chosen for high-temperature performance and resistance to a wide range of corrosive and high-temperature environments, including hydrogen-rich gas.

Dimensions: OD762 × 8 × 2007 mm, 987 kg per unit, two units.

The engineering challenges that special alloys exist to solve:

• High-temperature creep, oxidation and carburization. The tube material must resist creep and metallurgical degradation in a hydrogen-rich, >800 °C effluent over the design life. This is the single reason Incoloy/Inconel-class alloys are specified — and the single reason the vessel costs what it does. There is no cheap substitute that survives the duty.
• Differential thermal expansion. A large temperature gap between the hot process gas (tube side) and boiling water (shell side) creates significant differential expansion between tubes and shell. The design must absorb that movement — through expansion-joint or floating-arrangement provisions — without overstressing tube-to-tubesheet joints. Get this wrong and the failure shows up as tube-to-tubesheet leaks under thermal cycling.
• Welding special alloys is its own discipline. N08330 and Alloy 600 demand qualified procedures (WPS/PQR), tight heat-input control, the right filler metals, and careful PMI at every joint. This is not SA-516 welding with a different rod; it is a separate qualification regime. (Earlier Korea steam generators in the same family used SA-213 TP304H tubes on an SA-516 shell with SS-overlay tubesheet — the alloy choice tracks the exact temperature and environment of each duty.)

When a buyer asks "can your shop handle special materials," the E2151 is the answer in physical form: shell and tubes in two different high-nickel alloys, fabricated and PMI-traced to ASME U and KGS.

Vessel Deep-Dive 4: The Ultra-Large Chiller Water Tower (1FE-C171) — 29.5 Tonnes of Cooling

Not every challenge is metallurgical. The 1FE-C171 chiller water tower / direct-contact after-cooler from the Ulsan 6 A1700 project is an engineering problem of scale: DN4500 diameter, 29,550 kg — nearly 30 tonnes — in SA-516 Gr.70 with SS304 internals, SS316L packing supports and SA-350 LF2 nozzle forgings.

Direct-contact cooling design. A direct-contact after-cooler cools hot process gas by bringing it into intimate contact with cooling water through internal distribution systems, packing or trays, and demister pads. The internal geometry has to enforce uniform gas-liquid contact while minimizing pressure drop — because every millibar of pressure drop across this vessel is a compressor-efficiency penalty downstream.

Large-diameter fabrication discipline. A DN4500, tall, internally complex vessel pushes shop capability:
- Maintaining shell roundness during welding of a 4.5-metre-diameter shell — out-of-roundness is both a code limit and a fit-up problem for the internals.
- Precise positioning of internal support rings for packing and trays across that span.
- Nozzle orientation tolerances so the vessel drops into the plant piping layout without field rework.

This is where an 8,000 m² workshop with the cranes, rolls and fit-up jigs to handle a 30-tonne vessel is not a marketing line — it is the gate that decides whether you can bid the job at all.

Vessel Deep-Dive 5: The Air Receivers (C921 ×2) — The Unsung Reliability Buffers

The two C921 air receivers (SA-516 Gr.70, DN457 × 6 mm, 155 kg, from the Samsung P / Samsung Korea project line) are the simplest vessels in the package — and that is exactly why they matter. Air receivers buffer instrument-air and process-air supply, smoothing compressor pulsation and providing ride-through capacity. In a semiconductor-feeding hydrogen plant, where supply reliability is as critical as gas purity, the humble air receiver is part of the plant's stability backbone. Simple geometry, but the same ASME U + KGS code discipline, the same hydrotest, the same data book. A fabricator that treats the "easy" vessels casually is a fabricator you will eventually hear back from.

Material Selection Logic: Why Each Alloy, and Why Not the Cheaper One

Across these five vessel families, four material classes appear. Each was chosen for a specific reason, and each rejection of a cheaper option is itself an engineering decision. This is the section to read if you are about to argue a material spec with a fabricator or an EPC.

SA-516 Gr.70 — the carbon-steel workhorse (C501, 1FE-C171, C921)

SA-516 Grade 70 is a carbon-steel plate specified for moderate- and lower-temperature pressure-vessel service with improved notch toughness. It shows up on the PSA adsorbers, the chiller water tower and the air receivers because it is the right answer when the service is not aggressively hot or corrosive: excellent weldability, well-characterized mechanical and fatigue properties, broad availability, and a cost basis that no stainless or alloy can match. Roughly speaking, SA-516 Gr.70 covers the majority of pressure-vessel applications in this kind of plant — and over-specifying stainless where SA-516 suffices is a classic way to add cost without adding capability. The engineering discipline is knowing where SA-516 stops being adequate.

SA-240 304L — solid stainless for catalyst and moderate-hot service (C301)

The shift reactors use solid SA-240 304L because the service is hot (~350 °C), wet, CO₂-rich and hydrogen-bearing, with an internal catalyst bed. The "L" (low-carbon) grade matters: it limits carbide precipitation (sensitization) in the heat-affected zones of welds, preserving corrosion resistance at the weld lines — important in a multi-nozzle, multi-internal vessel with substantial welding. Solid 304L, rather than 304L-clad carbon steel, eliminates the lining-weld leak path and the liner/backing-plate differential-expansion concern. On a small-diameter DN273 vessel the material cost penalty of solid versus clad is modest, and the reliability gain is real.

SB-536 N08330 (Incoloy 800H) + Alloy 600 (Inconel) — high-nickel for >800 °C (E2151)

This is where the money goes, and for good reason. The steam generator sees reformer-effluent gas above 800 °C on the tube side. At that temperature:

• Carbon steel is disqualified outright — it creeps and oxidizes catastrophically.
• Standard 300-series stainless loses creep strength and is vulnerable to carburization in the carbon-bearing, hydrogen-rich gas.
• Incoloy 800H (N08330) is purpose-built for this regime: a nickel-iron-chromium alloy with controlled grain size and carbon for elevated-temperature creep-rupture strength, plus oxidation and carburization resistance.
• Inconel / Alloy 600 on the tubes adds nickel-chromium high-temperature performance and broad corrosion resistance, including in hydrogen-rich environments.

When a buyer flinches at the alloy cost of a reformer-effluent steam generator, the honest answer is that there is no cheaper material that survives the duty. The alloy is the function. (In an earlier Korea steam generator of the same family, the duty was met with SA-213 TP304H tubes on an SA-516 shell with SS-overlay tubesheet — a lower-temperature point on the same design philosophy. The exact alloy tracks the exact temperature.)

The Nelson curve and high-temperature hydrogen service

One material principle deserves its own mention because it governs hydrogen plants specifically: high-temperature hydrogen attack (HTHA). Hydrogen at elevated temperature and partial pressure can diffuse into steel and react with carbides, forming methane internally and causing fissuring and embrittlement. The governing reference is API 941, whose "Nelson curves" plot safe operating limits of temperature and hydrogen partial pressure for various steels. Where a vessel operates above the carbon-steel Nelson limit, the design moves to Cr-Mo alloy steels (SA-387 Gr.11/22) or austenitic stainless, often with stainless cladding/overlay for corrosion protection — exactly the material logic seen in the broader Air Products hydrogen-package vessels (SMR-effluent and shift-converter service). For an EPC buyer, the takeaway is concrete: ask your fabricator how each hydrogen-service vessel was checked against API 941. A shop that can answer fluently is a shop that has built hydrogen plants.

Material	Class	Used On	Chosen Because	Rejected Cheaper Option
SA-516 Gr.70	Carbon steel plate	C501, 1FE-C171, C921	Weldable, tough, economical for moderate service	— (already the baseline)
SA-240 304L	Austenitic stainless	C301	Uniform corrosion resistance + weld-zone integrity in hot catalyst service	304L-clad SA-516 (leak-path / expansion risk)
SB-536 N08330 (Incoloy 800H)	Ni-Fe-Cr alloy	E2151 shell	Creep / oxidation / carburization resistance >800 °C	300-series stainless (creeps, carburizes)
Alloy 600 (Inconel)	Ni-Cr alloy	E2151 tubes	High-temp + hydrogen-environment corrosion resistance	Standard stainless (degrades at duty temp)

The Fabrication Sequence: From Plate to Data Book

Knowing the material list is one thing; knowing how a multi-vessel, multi-material, dual-code package actually moves through a shop is what separates a credible fabricator from a hopeful one. Here is the fabrication arc for a package like the Air Products Korea hydrogen vessels.

1. Design and code calculation. Each vessel's shell thickness, head geometry, nozzle reinforcement and (for PSA) fatigue analysis are calculated to ASME Section VIII Div.1, with the design package built to satisfy both the ASME Authorized Inspector and the KGS reviewer. Getting design conditions, material equivalences and nameplate marking aligned to both regimes up front is what avoids a mid-fabrication rework loop.

2. Material procurement and receipt inspection. Plate, tube, forgings and fasteners are procured against the spec, then verified on receipt against mill certificates with PMI on alloys. For the E2151's N08330 and Alloy 600, this is the step where a mixed or substituted heat would be caught — and must be.

3. Welding procedure qualification. Each material/joint combination needs a qualified WPS backed by a PQR. Carbon steel, 304L, and the high-nickel alloys each require separate qualification. A shop maintaining 600+ PQRs typically already has the relevant procedures on the shelf, which is the difference between starting fabrication this week versus qualifying procedures for a month first.

4. Forming and fit-up. Plates are rolled to shells, heads formed, nozzles and internals fitted. For the DN4500 chiller tower, maintaining shell roundness through this stage is a measurable code concern; for the 16 PSA vessels, jig consistency across the batch is what guarantees interchangeability.

5. Welding and in-process NDE. Production welding follows the qualified procedures with controlled heat input — especially critical for the alloy steam generator. NDE (RT/UT/PT/MT) is applied per code and per Air Products' enhanced coverage, documented on weld maps and coverage diagrams as the work proceeds, not reconstructed afterward.

6. Internals installation. Catalyst supports, bed grids, distributors, demisters and packing supports are installed to tolerance — the step that determines whether the shift reactors and PSA adsorbers actually perform their process duty rather than just holding pressure.

7. Hydrostatic test. Each completed vessel is hydrotested to its code test pressure with calibrated instrumentation and documented hold time — the witnessed proof-of-integrity milestone.

8. Surface preparation, painting and PWHT where required. Post-weld heat treatment is applied where the code, thickness or service demands it; surface protection is applied per the project's preservation spec for overseas transit.

9. Data-book assembly and final release. Every record — drawings, calcs, mill certs, PMI, weld maps, NDE reports, hydrotest charts, the ASME Manufacturer's Data Report and KGS documentation — is compiled into an auditable data book, and the vessel is released under U Stamp with KGS registration complete.

A buyer who understands this sequence can ask better questions at each gate — and a fabricator who has run it repeatedly for Air Products can answer them without hesitation.

Packing, Preservation and Shipping to Korea

The vessel that passes hydrotest in the shop is not the vessel that matters — the vessel that arrives in Korea undamaged and rust-free is. For a hydrogen-plant package shipping by sea to a Korean port, preservation and packing are part of the engineering, not an afterthought.

• Moisture and corrosion control. Internals and nozzle faces are protected against the humidity of a multi-week ocean voyage; sealed nozzle covers, desiccant and VCI (vapor-corrosion-inhibitor) measures keep machined and stainless surfaces clean.
• Shock and securing. Large items like the ~30-tonne 1FE-C171 tower require engineered cradles, lashing and lifting-point design so the vessel survives handling and sea motion without distortion to shell roundness or internals.
• Documentation travels with the steel. The data book and code documentation are prepared so customs and the receiving KGS inspection have what they need on arrival, avoiding port delays.

This is mundane until it goes wrong — a rust-streaked stainless internal or a tweaked nozzle flange at site is an expensive, schedule-eating problem. A fabricator with experience exporting pressure equipment to 50+ countries treats packing as a deliverable in its own right.

The Dual-Code Reality: ASME U and Korean KGS

Here is the part that catches EPC teams off guard when they first export pressure equipment into Korea: ASME U Stamp alone does not clear customs into a Korean plant. Korea regulates pressure equipment under its own Pressure Equipment Safety / KGS (Korea Gas Safety) code regime, and equipment must be registered and inspected accordingly (KGS / Korea hydrogen regulation overview, CMS Expert Guides). For an Air Products Korea hydrogen package, that means every vessel carries two code obligations simultaneously:

1. ASME Section VIII Division 1, U Stamp — design, materials, fabrication, NDE, hydrotest and the Manufacturer's Data Report under an ASME Certificate of Authorization, witnessed by an Authorized Inspector (AI).
2. KGS registration — Korea-specific design review, documentation and inspection requirements layered on top, so the vessel is legally accepted for installation and operation in Korea.

The practical implications for a fabricator:

• Design documentation has to satisfy two reviewers. Calculations, drawings and material certs are prepared so they pass both the ASME AI and the KGS review — alignment of design conditions, material equivalences and nameplate marking up front, not as a rework loop.
• Material traceability is doubled-down. Both regimes want full mill-cert traceability and PMI. For the high-nickel alloys in E2151 (N08330, Alloy 600), traceability is not negotiable.
• Inspection and hold points are coordinated so a single fabrication sequence satisfies ASME hold points and any KGS witness requirements — rather than stopping the line twice.

A fabricator that has already run this dual-code gauntlet for Korea — repeatedly, for a client as demanding as Air Products — removes a category of schedule and compliance risk that a first-timer would discover the hard way, mid-fabrication.

A note on the "blue hydrogen" question. Because Korea's CHPS and clean-hydrogen auctions are reshaping which projects get funded (InvestKorea), buyers sometimes assume a "clean hydrogen" plant needs fundamentally different vessels. It does not. A blue-hydrogen SMR plant is a grey-hydrogen SMR plant plus a CO₂-capture and compression block downstream. The shift reactors, PSA adsorbers, steam generators and after-coolers are the same vessel families — which is precisely why this Air Products reference list applies directly to the next wave of carbon-capture-equipped hydrogen projects.

The Korean Hydrogen Market Context — and Why the Vessel Mix Is Stable

It is worth situating this project inside the Korean market it serves, because the policy turbulence of 2025–2026 has a counterintuitive implication for vessel procurement.

Korea has been building a hydrogen economy through the Clean Hydrogen Portfolio Standards (CHPS), a mechanism that incentivizes power producers to buy clean hydrogen via long-term contracts, with a forward market intended to open in 2027 (Baker McKenzie hydrogen tracker). The 2025 clean-hydrogen auction added an exchange-rate-linked settlement and a volume-borrowing mechanism to manage currency risk for developers (InvestKorea). But in June 2026, the Ministry of Climate, Energy and Environment cut the general Hydrogen Power Bidding volume nearly 30% and shrank the CHPS auction to roughly one-sixth of its earlier level, and excluded coal-ammonia co-firing from the clean market (Seoul Economic Daily, June 2026). Under the current administration, hydrogen remains designated a national strategic industry eligible for new tax incentives.

So the volumes and timing of Korean hydrogen-for-power are in flux. What is not in flux is the demand from Korea's industrial base — the semiconductor fabs, petrochemical complexes at Ulsan, and steel and chemical plants — for merchant hydrogen and industrial gases supplied by majors like Air Products. That demand is structural: a chip fab needs ultra-pure hydrogen whether or not a hydrogen-power auction clears this quarter. And merchant hydrogen for those customers is made by exactly the SMR–shift–PSA train this article describes.

The procurement implication is clean: a fabricator's exposure to Korean hydrogen-vessel demand is far more stable than the auction headlines suggest. The vessels feeding Samsung's semiconductor complex (the C921 air receivers, the Samsung P package) and the Ulsan air-separation and gas plants (the 1FE-C171 cooling tower) are driven by industrial production, not by power-auction volumes. When you read that Korea cut its hydrogen auction by 30%, do not conclude that hydrogen-plant vessel demand collapsed — the industrial-gas backbone that Air Products serves is a different, sturdier market.

Air Products' Broader Hydrogen Vessel Portfolio — Context for This Package

The five vessel families in the headline list sit within a wider body of Air Products work that Lmart has fabricated, and seeing the full set helps an EPC buyer understand the depth of capability behind a single package:

• Desulfurization & heat-recovery vessels — ASME U, shell SA-516 Gr.70 with TP304H internal lining, SS316L catalyst supports; high-temperature (350–450 °C) sulfur-bearing service with continuously welded, leak-tested internal lining (delivered 2025).
• Molecular-sieve adsorbers, demisters and activated-carbon filters — for Air Products' Mubea CZ HPPSA63-180-3 high-purity PSA hydrogen plant; ASME U, multiple distinct vessel types in one project scope, with the tight internal tolerances that PSA-class separation demands (delivered 2025).
• Carbon-adsorption towers — ASME U, SA-516 Gr.70 shell with TP304 internals and SS316L bed screens; nozzle-load calculations per ASME and WRC 537, with Air Products' documentation requirements exceeding standard ASME minimums (delivered 2025).
• Hydrogen plant equipment package (Samsung P) — multiple ASME-coded vessels including SMR-effluent vessels, PSA vessels, shift converters and separators; high-temperature components in SA-387 Gr.11/22 Cr-Mo steel with TP347 SS overlay, designed against API 941 for hydrogen service (delivered 2026).
• Chiller water towers & direct-contact after-coolers (Ulsan 6 A1700) — large-diameter cooling vessels with engineered gas-liquid contact internals (delivered 2026).
• High-temperature waste-heat steam generators (Korea, 2021) — the E2151 family, recovering reformer-effluent heat into HP steam.

The pattern across all of it is consistent: carbon steel where the service allows, stainless and high-nickel alloys where the temperature and hydrogen environment demand, documentation that exceeds the code minimum, and dual-code compliance for the Korean destination. A single package is impressive; a multi-year, multi-project record across a client's most demanding qualification process is what actually de-risks an EPC's award decision.

Five Procurement Mistakes That Sink Hydrogen-Vessel Packages

Drawing the engineering lessons together, here are the recurring mistakes that turn a hydrogen-vessel package into a schedule and cost problem — and how the Air Products Korea approach avoids each.

Mistake 1 — Treating a PSA vessel like a static drum. PSA is fatigue service. Designing only to static pressure, with sharp nozzle transitions and partial-penetration internal welds, sets up cyclic cracking. Avoid by: explicit ASME VIII cyclic/fatigue design, reinforced nozzles, smooth transitions, full-penetration internal attachments — across all 16 vessels identically.

Mistake 2 — Under-specifying steam-generator alloys to save cost. Substituting standard stainless for Incoloy/Inconel in >800 °C reformer-effluent service is a false economy that ends in creep failure and tube-to-tubesheet leaks. Avoid by: matching the alloy to the actual temperature and environment, and accepting that the alloy is the function.

Mistake 3 — Ignoring API 941 / HTHA in hydrogen service. High-temperature hydrogen attack is invisible until it isn't. Avoid by: checking every hydrogen-service vessel against the Nelson curve and moving to Cr-Mo (SA-387) or clad/overlay construction where required.

Mistake 4 — Discovering KGS at the export stage. A vessel built to ASME U alone, then found to need KGS registration after fabrication, means rework, re-documentation and delay. Avoid by: designing to both codes from day one, with aligned design conditions, material equivalences and inspection hold points.

Mistake 5 — Treating internals as secondary. Bed supports, distributors and demisters built loosely cause channeling, hot spots and premature catalyst/adsorbent failure — process problems that surface months after delivery and are blamed on the catalyst, not the vessel. Avoid by: fabricating internals to the same tolerance discipline as the pressure boundary, with documented jigging.

Every one of these mistakes is cheaper to avoid in the shop than to discover in the field. The reason a proven reference list matters is that the proof is the avoidance.

How to Evaluate a Fabricator for an SMR / Blue-Hydrogen Vessel Package

If you are an EPC procurement engineer or process lead qualifying a shop for a hydrogen-plant package, here is the decision framework this project illustrates. Use it as a checklist.

1. Demand a delivered hydrogen-vessel reference list — not a capability brochure

Anyone can list "PSA vessels" and "shift reactors" on a website. Ask for the as-built record: tags, quantities, materials, dimensions, weights, year delivered, and the end client. A fabricator who has actually built sixteen C501 PSA adsorbers and two C301 shift reactors for a named industrial-gas major has solved problems a first-timer hasn't met yet.

2. Test for special-alloy capability specifically

Carbon-steel SA-516 fabrication is table stakes. The differentiator is whether the shop can weld and PMI-trace Incoloy 800H (N08330) and Inconel (Alloy 600) to code, with qualified WPS/PQR and the heat-input control those alloys demand. The E2151 steam generator is the litmus test: if a shop can deliver that, it can handle the alloy work your hydrogen plant will eventually need.

3. Verify cyclic / fatigue design competence for PSA service

PSA is a fatigue service. Ask how the fabricator handles nozzle reinforcement, weld-profile control and internal-attachment design for cyclic loading. A shop that talks only about static design pressure has not internalized why PSA vessels fail.

4. Confirm internals capability, not just the pressure boundary

For shift reactors and PSA/adsorber vessels, the internals determine process performance: bed supports, hold-down grids, distributors, demisters. Channeling and maldistribution are fabrication-tolerance problems that masquerade as process problems. Ask to see the internal jigging and tolerance approach.

5. Confirm large-vessel shop capacity

A DN4500, 30-tonne vessel like 1FE-C171 needs the cranes, rolls and floor space to fabricate it, plus the heavy-lift and shipping plan to move it. Confirm the workshop footprint (8,000 m²+) and lifting capacity match your largest single item — before award, not after.

6. Confirm the export-code path you actually need

For Korea, that means ASME U + KGS, run in parallel. For Europe it means PED/CE; for other markets, the relevant classification or national code. A fabricator with 600+ PQR welding qualifications, the ASME U-Stamp, PED, and delivery into 50+ countries has the documentary infrastructure to clear your destination's code — ask them to name the exact regime for your project and describe how they'll satisfy it.

7. Inspect the documentation discipline

Air Products requires manufacturing documentation exceeding standard ASME minimums — weld maps, NDE coverage diagrams, PMI records, full traceability data books. The quality of a fabricator's paperwork is a leading indicator of the quality of their welds. Ask to see a sample data book from a comparable project.

Quality, Inspection and the Data Book

Behind every vessel in the table is an inspection and test plan (ITP) that turns code requirements into a sequence of witnessed and held inspection points. For an Air Products Korea hydrogen package, the quality spine looks like this:

• Material receipt and PMI. Every plate, tube and forging verified against mill certs; positive material identification on alloys — critical for the N08330 / Alloy 600 E2151 and the 304L C301, where a mixed heat would be catastrophic.
• Weld procedure qualification. WPS/PQR qualified for each material and joint configuration — carbon steel, 304L, and the high-nickel alloys each demand their own qualified procedures. (Lmart maintains 600+ PQR welding qualifications precisely so multi-material packages like this don't stall on procedure development.)
• NDE coverage. RT/UT/PT/MT per code and per Air Products' enhanced requirements, documented on coverage diagrams so the AI and KGS reviewer can see exactly what was examined.
• Hydrostatic testing. Each vessel hydrotested to code, with calibrated gauges and documented hold times — the proof-of-integrity milestone before final release.
• The data book. A complete manufacturer's record — drawings, calcs, mill certs, PMI, weld maps, NDE reports, hydrotest records, the ASME Manufacturer's Data Report and KGS documentation — assembled so the buyer receives a single auditable package per vessel.

This documentation discipline is not bureaucracy; it is the mechanism by which a buyer thousands of kilometres away can trust a vessel they never watched being built. For Air Products — whose vendor-qualification process is among the most demanding in the industrial-gas sector — it is the price of admission, and it is the standard we build to as a matter of routine.

Lmart Capability — Factual Positioning

Lmart (Suzhou Lmart Energy Equipment Co., Ltd., a member of Huachang Group) is an ASME-certified pressure-equipment manufacturer with the ASME U-Stamp, PED 2014/68/EU (CE marking), and ISO 9001 / 14001 / 45001 systems. We hold 600+ PQR welding qualifications, run a full in-house workflow from design through fabrication, assembly, testing and documentation in an 8,000 m² integrated workshop, and have delivered pressure equipment, heat exchangers, refrigeration and gas-compression packages, and modular skids into 50+ countries.

The Air Products Korea hydrogen record described here — PSA adsorbers, shift reactors, special-alloy steam generators, ultra-large coolers and air receivers, all to ASME U + KGS — is one slice of a hydrogen and industrial-gas portfolio that also includes desulfurization and heat-recovery vessels, molecular-sieve adsorbers, carbon-adsorption towers, and high-pressure hydrogen storage tanks. We mention it not as a sales claim but as evidence: when an EPC team needs a coordinated, multi-material, dual-code hydrogen-plant vessel package delivered to a demanding industrial-gas client, this is work we have done before and are equipped to do again.

Conclusion

Hydrogen's economics are being rewritten in real time — Air Products' Louisiana FID targeted for mid-2026, Rotterdam carbon capture coming online, Korea's CHPS auction resized. But the hardware that makes hydrogen — shift reactors, PSA adsorbers, waste-heat steam generators, after-coolers — sits underneath all of it, indifferent to the policy label on the molecule. That is the quiet advantage of a proven vessel reference list: it is durable in exactly the way the market is not. For an EPC buyer specifying an SMR or blue-hydrogen package bound for Korea, the questions that matter are concrete and answerable — Can you build PSA vessels for fatigue service? Can you weld Incoloy and Inconel to code? Can you carry it through ASME U and KGS? Have you done it for a client as demanding as Air Products? On this project, the answers are in the table.

Frequently Asked Questions

Q1: What is a PSA vessel and why does an SMR hydrogen plant need so many of them?
A PSA (pressure-swing-adsorption) vessel is an adsorber that purifies the shifted syngas to high-purity hydrogen — typically 99.99%+, often 99.999%+ — by trapping CO₂, CO, CH₄ and trace gases on an adsorbent bed under pressure, then regenerating by depressurization. Because each vessel must cycle between adsorption and regeneration, a plant needs several adsorbers on a staggered schedule to keep producing continuously while individual beds regenerate. The Air Products Korea package here uses sixteen PSA vessels (tag C501, SA-516 Gr.70, DN273 × 9.27, 207 kg each), designed for the cyclic fatigue loading that defines PSA service.

Q2: Why is the shift reactor built from solid 304L instead of clad carbon steel?
The water-gas-shift reactor operates hot (~350 °C in this duty) in a wet, CO₂-rich, hydrogen-bearing environment with an internal catalyst bed. Solid SA-240 304L gives uniform corrosion resistance on every wetted surface and full PMI traceability without the leak-path and differential-expansion risks of a lining or overlay. For a small-diameter vessel like DN273, solid stainless is often the cleaner, lower-risk choice than clad SA-516 — no lining welds to leak-test, no liner/backing-plate expansion mismatch.

Q3: Why does the steam generator use Incoloy 800H and Inconel instead of stainless steel?
The E2151 steam generator recovers reformer-effluent heat with tube-side process gas exceeding 800 °C. At those sustained temperatures, ordinary stainless steels creep and degrade. Incoloy 800H (SB-536 N08330) on the shell side provides high-temperature strength plus oxidation and carburization resistance; Inconel (Alloy 600) on the tube side handles the high-temperature, hydrogen-rich gas. These high-nickel alloys are specified precisely because there is no low-cost substitute that survives the duty over the design life.

Q4: Does ASME U Stamp certification mean a vessel can be installed in Korea?
No. Korea regulates pressure equipment under its own KGS (Korea Gas Safety) / Pressure Equipment Safety code regime. Equipment exported into a Korean plant must satisfy both ASME Section VIII Div.1 (U Stamp) and KGS registration and inspection. A fabricator builds the design documentation, material traceability and inspection sequence to clear both reviewers in parallel — which is a meaningful schedule and compliance risk a first-time-into-Korea shop tends to underestimate.

Q5: Are the vessels for a "blue hydrogen" plant different from a conventional grey-hydrogen plant?
Largely no. A blue-hydrogen SMR plant is a grey-hydrogen SMR plant with a CO₂-capture and compression block added downstream. The core vessels — shift reactors, PSA adsorbers, waste-heat steam generators, after-coolers — are the same vessel families. That is why a delivered reference list of grey-hydrogen-plant vessels applies directly to the next wave of carbon-capture-equipped (blue) hydrogen projects, including those responding to Korea's CHPS clean-hydrogen framework.

Q6: What standards and codes govern these hydrogen-plant vessels?
Design and fabrication follow ASME Section VIII Division 1 (with U Stamp under an ASME Certificate of Authorization, witnessed by an Authorized Inspector), plus KGS registration for Korea. Materials follow their respective ASME/ASTM specifications — SA-516 Gr.70 carbon-steel plate, SA-240 304L austenitic stainless, SB-536 N08330 (Incoloy 800H) and Alloy 600 (Inconel) high-nickel alloys — with full mill-certificate traceability and PMI. High-temperature hydrogen-service material selection additionally references API 941 (Nelson curve) where Cr-Mo or high-nickel alloys are required.

Q7: How do you evaluate whether a fabricator can handle a multi-vessel hydrogen package?
Demand a delivered (as-built) reference list with tags, quantities, materials and weights; test specifically for special-alloy (Incoloy/Inconel) welding and PMI capability; verify cyclic/fatigue design competence for PSA service; confirm internals (bed support, distributor, demister) fabrication tolerance; confirm shop capacity for your largest single vessel (e.g. a DN4500, ~30-tonne tower needs the cranes and floor space); confirm the exact export-code path for your destination (ASME U + KGS for Korea); and inspect a sample data book — documentation quality is a leading indicator of weld quality.

Q8: Why does a single PSA hydrogen plant need sixteen adsorber vessels rather than two or three?
PSA produces hydrogen continuously even though each individual bed must periodically stop adsorbing to regenerate. To keep product flowing without interruption, the beds run on a staggered cycle — while some adsorb at high pressure, others are depressurizing, purging, or re-pressurizing. A larger number of beds enables a smoother, higher-recovery cycle and steadier product purity. The exact bed count is set by the PSA cycle design, plant capacity and target recovery; in this Air Products package the count is sixteen (C501). For a fabricator, the practical consequence is a serial build where all sixteen vessels must be dimensionally and metallurgically interchangeable.

Q9: What is the difference between high-temperature and low-temperature shift, and does it change the vessel?
High-temperature shift (≈350–450 °C, iron-chromium catalyst) reacts fast and removes the bulk of the CO; low-temperature shift (≈200–250 °C, copper-zinc catalyst) then pushes the equilibrium further toward hydrogen to minimize residual CO. Many plants run both in series, which is why this package contains two shift reactors (C301). The two stages share the same vessel family and 304L material logic, but differ in design temperature and catalyst, so design conditions, bed geometry and any internal heat-management features are set per stage. From a procurement standpoint, "two shift reactors" usually signals a properly staged shift section, not redundancy.

Looking Ahead

The hydrogen plants being sanctioned now — Air Products' blue-hydrogen developments, carbon-capture retrofits at existing grey plants, and the industrial-gas installations feeding Korea's fabs and petrochemical complexes — will all be built around the same reformer, shift and PSA core, with capture and compression layered on for the low-carbon variants. That means the vessel families documented here are not a snapshot of one finished project; they are the recurring building blocks of the hydrogen plants still to come. For the EPC teams specifying them, the most valuable thing a fabricator can show is not a forecast but a record: vessels delivered, materials traced, codes cleared, client satisfied. On the Air Products Korea hydrogen projects, that record is what the table at the top of this article represents — and it is the basis on which the next package gets built.

Related Reading

• Hydrogen Economy & Pressure Vessel Demand 2026
• Special-Alloy Vessels for Corrosive Chemical Service
• Pressure Vessel Market 2026: 4 Growth Pillars
• EPC Procurement Hub — pillar guide for buyers

---

Author: Qiangbin Chu, Suzhou Lmart Energy Equipment Co., Ltd. (member of Huachang Group). ASME U-Stamp · PED 2014/68/EU · ISO 9001/14001/45001 · 600+ PQR · delivered to 50+ countries.