Facts About Underwater Welding: Methods, Safety, Codes, and Applications

A damaged offshore pipeline begins leaking far below the surface, and a ship hull needs a critical patch before it can reach dry dock. In those high-stakes moments, only a skilled diver-welder can make a permanent or salvage repair without pulling the entire structure out of the water.

Introduction: Why Underwater Welding Matters

Underwater welding joins metals while the arc and weld pool sit beneath the surface. That plain description hides a lot of engineering: water changes how heat leaves the weld and how the electrode coating behaves, and it puts the welder inside a steel box a diver must climb into. What looks like a niche trade is really a critical enabler for marine construction, offshore energy production, port infrastructure, and salvage work worldwide.

The technique exists for economic reasons, and the numbers are stark. The alternative to welding a damaged pipeline section 30 meters down is isolating it, dewatering it, cutting it out, and fitting a replacement spool while the platform above keeps running. For a ship it means dry docking: find a berth, move the vessel, block it up, and wait. For a lock gate or an intake screen, it means shutting the facility down. Each time, the underwater weld turns a multi-week, multi-million-dollar shutdown into a job a small crew can finish in a shift or two.

The quality bar climbs steeply between the two. A temporary patch may only need to hold pressure through the next storm; a permanent repair must be designed around the loads, the environment, and the fitness of the base metal, then documented against a recognized standard. AWS D3.6M, the American Welding Society’s code for underwater welding, organizes the work into classes that reflect this question, with Class A for the most demanding applications and Class O for base metals whose weldability has not been established. Welds qualified to that level count as permanent structural elements, not field fixes. Procedure qualification and welder performance sit in ISO 15614-9, which covers testing for underwater wet welding procedures.

Where the work actually happens

Construction, repair, and salvage across a wide spread of marine and offshore assets:

Sector	Typical underwater welding work
Offshore oil and gas	Pipeline repair, riser and clamp connections, platform leg repairs, subsea tie-ins
Marine construction and shipbuilding	Hull section assembly in flooded graving docks, structural steel connections
Ship repair and maintenance	Patch plating, crack repairs, propeller and rudder work, thruster repairs
Ports and civil works	Quay walls, sheet piling, lock gates, pump station intakes, outfalls
Power, nuclear and decommissioning	Intake structures, reactor component work, spent fuel pool repairs
Salvage	Wreck stabilization, plugging and patching of hulls before recovery

Ship repair is where ordinary shop rules stop applying. Shop welders on hull steel usually handle carbon steel and aluminum plate with well understood procedures, as covered in AJ Marine Welding: Processes, Standards, Safety and Best Practices. Underwater the same base metals behave differently, access is worse, and the workable window is hours, not days.

The two skills that have to exist at the same time

The biggest constraint on underwater welding is not metallurgy, it is the diver. A welder underwater performs two trades at once, commercial diving and welding: managing depth, breathing gas, decompression obligations, entanglement hazards, current, and a hard limit on bottom time, while holding arc control in near-zero visibility, where the puddle shows only when the bubble clears. No shop-floor skill substitutes for that training, and no diving skill substitutes for knowing how a coating melts.

That dual competency drives everything else. Most work is done by two people, a diver-welder paired with a tender who manages surface supply and monitors the job, and surface support, current monitoring, and the position of the work are planned as carefully as the amperage. That is also why shielded metal arc welding dominates the field. Why SMAW still wins is well documented, and underwater the reasons are amplified: the process is simple, the equipment is rugged, and the flux shield covers the molten metal even as the water tries to chill it.

Preparation also differs. Cutting and gouging steel underwater to clean up a damaged joint is a separate discipline with its own hazards, covered in Air Arc Welding and Carbon Arc Gouging, and it often takes as much time as the welding itself.

Why the field still moves forward

Underwater welding is mature in some respects and unresolved in others. The codes are established and the committees that maintain them are active, with the AWS D3 Committee on Welding in Marine Construction refining practice for the marine environment. Recent academic work in journals such as Welding in the World keeps pushing on rapid cooling, hydrogen pickup, and the brittle microstructures the underwater environment creates.

This article works through those layers in order: the three core methods, wet, dry hyperbaric, and hybrid; processes, equipment, and consumables; the metallurgy behind porosity and hydrogen cracking; the codes and classification society expectations; applications, benefits, and limits; and finally safety practice, recurring mistakes, and troubleshooting.

For anyone working in the marine environment, or managing structural repairs where dry access is not an option, the goal of the sections that follow is simple: to make the decisions underwater welding demands, and to make them safely.

History and Evolution of Underwater Welding

Underwater welding did not start as an engineered process. It started as emergency improvisation, by people who needed a hull mended before a tide or a storm window ran out. Nearly every rule the industry follows today, from electrode choice to inspection hold points, traces back to something that went wrong in a diving helmet in the 1940s.

Electric arc welding was commercially available from the early 1900s, so the arc itself was never the hard part. The hard parts were everything around it: breathing gas, darkness, cold, water pressing into every gap, and molten steel meeting seawater instead of air.

Salvage Roots: Welding from Diving Suits

Early attempts date to the first decades of the 20th century, when salvors and shipyard crews improvised repairs in shallow water using hard-hat diving gear. Results were rough: porous weld metal, brittle fusion zones, and no accepted way to prove a joint was fit for structural duty.

During and after the Second World War, wet welding hardened into a recognized salvage specialty. Divers in surface-supplied helmets relied on the arc for illumination in near-zero visibility and welded mainly small, shallow beads. The unspoken rule: keep it small and accept it as temporary. No one treated a wet weld as a shop weld, and for decades the process sat outside the welding codes with no way to test it.

The Consumables Breakthrough

The turning point came less from the power source than from the electrode. Through the 1950s and 1960s, suppliers developed wet-welding electrodes, including the Alrode process used for offshore pipeline splash-zone repairs and the Ozarc electrodes developed in the United Kingdom with the British Welding Research Association. Their rutile-type, oxidizing coatings burned off steadily underwater, held the arc stable in moving water, and produced a slag that shielded the pool from the sea.

The dry branch ran on a different track. Hyperbaric chambers and habitats kept welders dry at shallow depth from the 1950s, and by the 1970s saturation techniques allowed dry welding much deeper. Research mapped the wet process as well, including the faster cooling water imposes on the weld pool and the hydrogen a wet electrode introduces, which is why heat input control matters so much.

The manual arc process remained the workhorse, a conclusion that still holds in the broader case for stick welding. Purpose-made consumables and better chamber logistics turned an improvised patch into something specifiable.

Building the Code: AWS D3.6 and the Classification Era

By the early 1970s, field data made it clear that underwater welding ought to be written down. The American Welding Society’s D3 Committee on Welding in Marine Construction formed a Subcommittee on Underwater Welding in 1975, and the first AWS D3.6 was published in 1983. Later revisions, including the current AWS D3.6M, refined testing, documentation, and quality levels.

The standard reorganized the practice around graded quality. Class A is the most stringent, requiring the widest range of mechanical testing and the tightest acceptance limits. Class B is intermediate, and Class O covers other, less critical work. A repair is no longer just called underwater welding; it is a Class A, B, or O weld, and the class sets the testing and documentation that follow.

Internationally, ISO 15614-9, part of the ISO 15614 qualification series, covers procedure qualification for hyperbaric wet welding. Electrode selection, including all-position classifications, forms part of that package, as shown in our all-position welding rod guide.

Classification societies formed the other half of the story. ABS, DNV, and Lloyd’s Register historically refused to accept underwater welds on faith, requiring testing and review of the procedure before approval. That pushed the industry toward documented, repeatable work, frameworks ship repair contractors still work inside, as outlined in our AJ Marine Welding processes and standards overview.

Four Decades of Research

Since the 1980s, research has turned wet welding into a code-qualifiable process in shallow water, refining consumable chemistry, quantifying hydrogen pickup, optimizing heat input, and validating mechanical properties against AWS and ISO acceptance criteria. The U.S. Department of Energy’s technical bibliography on underwater welding charts much of that literature, and recent peer-reviewed reviews in Welding in the World show where the practice stands.

Three shifts define the result:

• Salvage to engineered repair. An underwater weld can now be specified as a permanent structural repair in defined classes.
• Uncontrolled to documented. Consumable storage, procedure qualification, welder continuity, and nondestructive examination are all part of the record.
• Manual to assisted. ROV and diver-assisted systems have moved from research curiosity toward limited production use, with depth the remaining constraint.

Period	Milestone	What Changed
1900s to 1930s	Improvised shallow-water wet welding	Porous, brittle results
1940s to 1950s	Salvage trade formalized	Temporary repairs, no qualification
1950s to 1970s	Wet consumables, hyperbaric chambers	Stable arc, dry repairs at depth
1975	AWS D3 forms Underwater Welding Subcommittee	Standardization begins
1983	First AWS D3.6 published	Code framework for underwater welds
1980s to 1990s	Hyperbaric qualification research	Tested, repeatable procedures
2010s to today	D3.6M revisions, ISO 15614-9, ROV trials	Graded classes, first automation

That evolution, from salvage improvisation to graded, code-driven repair work, sets up the methods, consumables, and safety rules that follow.

Underwater Welding Methods: Wet, Dry (Hyperbaric), and Hybrid

Underwater repair work falls into three method families, and the choice is driven by access, geometry, depth, base metal condition, and the quality the structure needs. A splash-zone grating repair and a deep-water pipeline girth weld pose very different problems. Get it wrong and you burn hours, money, and consumables on a repair that cannot be code-accepted.

Wet Welding: Doing the Job in Open Water

Wet welding is exactly what the name suggests. There is no mechanical barrier between the arc and the water: the welder, electrode, and workpiece are submerged, and the arc burns directly in the sea. Only a bubble of steam and gas around the electrode tip, formed as the flux decomposes, keeps the molten pool from being quenched instantly. It is a rough, unstable shield, but it handles much underwater repair work.

Shielded metal arc welding is the traditional workhorse, and stick process has survived many changes in general fabrication for a reason: its forgiving arc, simple electrode handling, and tolerance for dirty surfaces matter underwater, where cleaning and joint access are limited. Flux-cored arc welding has gained ground because it deposits more metal per hour and can produce a more ductile, crack-resistant bead. Recent peer-reviewed reviews of underwater welding show how electrode chemistry changes deposit chemistry in the quenched, hydrogen-rich wet environment.

The tradeoffs are clear: wet welding is the fastest, cheapest, and most flexible option, and with no chamber to fabricate, position, and dewater, the job starts sooner. Depth is the biggest limit, with wet welds generally restricted to around 30 m (100 ft) and most quality work inside 10 m (30 ft). Visibility, current, and bottom time compete with arc time, and the bead comes out fast-cooled, porous, and often hard and brittle. That makes wet welding a strong candidate for temporary or non-critical repairs and a poor candidate for a pressure-retaining joint that must pass inspection.

Dry Hyperbaric Welding: Chambers, Habitats, and Bells

Dry hyperbaric welding moves the arc out of the water and into a gas-filled space at or near ambient pressure: a surface-supported welded habitat lowered over the work area, a rigid steel chamber, or a closed diving bell. The welder works in air or a breathing mix, the joint stays dry, and the arc behaves as it does in a shop. The result is a weld that looks and tests like a surface weld, and AWS D3.6M sets the framework for wet and hyperbaric underwater welding, covering inspection, personnel, and habitat requirements.

The catch is everything around the weld. A chamber has to be designed, transported, positioned, sealed, and filled with a breathable gas mix that avoids nitrogen narcosis at working depth, then reversed when the job ends. That makes dry hyperbaric welding expensive and slow, and the hardest constraint is geometry: anything larger than the chamber interior cannot be welded in it. In shallow water, surface habitats are the practical choice; deeper work pushes to surface-supplied habitats and closed bells, where gas management gets more demanding. The AWS D3 Committee on Welding in Marine Construction maintains this qualification guidance.

Hybrid Systems: Splitting the Diver and the Workpiece

Hybrid systems divide the work between environments. The most common pattern is a dry chamber or gas-filled bell enclosing only the weld joint, with the diver outside in the water feeding the electrode or positioning the torch through an access port. Another keeps fit-up and grinding dry inside a habitat, then moves the final weld outside. It borrows dry welding’s quality control without a full chamber around the whole structure, though it needs specialized equipment and costs more than a straight wet weld. Some crews also work in a cofferdam or seawater-exclusion dome, flooding the work area to ambient pressure and welding in a dry, open space, which suits surface and near-surface work where the water can be pumped away.

Choosing the Right Method

Factor	Wet	Dry Hyperbaric	Hybrid
Arc environment	Arc burns in open water, no barrier	Gas-filled chamber, near-ambient pressure	Dry joint, wet surroundings
Processes	SMAW, increasingly FCAW	GTAW, SMAW, other standard processes	SMAW or FCAW in chamber, wet prep outside
Depth	About 30 m (100 ft); quality work shallower	Habitat design; bells and habitats go deeper	Similar to dry methods
Setup and cost	Lowest cost, fastest mobilization	Highest cost, longest mobilization	Mid-range
Weld quality	Fast-cooled, porous, less predictable	Comparable to surface welding	Approaches dry quality
Best suited for	Salvage and temporary repairs	Code-critical joints, new fabrication	Small joints, awkward access

Consumables matter as much as the environment. Wet welding needs electrodes and wire that tolerate rapid quenching and hydrogen pickup, and damp electrodes fail fast in storage. Dry methods run conventional products with familiar shop-style shielding gas. Why SMAW still wins covers the process fundamentals that carry over, and Airgas welding supply buying guidance covers the gas and cylinder side.

Note that ISO 15614-9 covers qualification testing of wet welding procedures for hyperbaric application, and qualifying a procedure is separate from welding the joint. The AWS marine cutting and welding guidance is a good reference for how these repairs are specified and accepted.

Processes, Equipment, and Consumables

Most underwater welds in repair reports came from a process invented in the 1880s. Shielded metal arc welding, or stick welding, still does the bulk of open-water work: no shielding gas, no cooling loop, little support equipment. Everything else is a variation on that theme, which is why the industry has stayed conservative.

Why SMAW Dominates Wet Welding

SMAW is the default wet process because the arc sits inside a flux coating that decomposes into a shielding gas and slag. That self-shielding action makes the process forgiving in a hostile environment. Open water is conductive, so the diver cannot blanket the work area in argon; the coating becomes the shield and works only when formulated for the job.

The electrode is the consumable. A diver holds a holder on a stinger, feeds the rod as it burns, and manages the arc by hand while controlling body position in the water. The process is slower than its dry-land cousins and the arc is a constant source of bubbles and heat, but it is predictable and backed by decades of qualification data. AWS D3.6M, the standard for underwater welding, was built around processes like this.

Modern wet practice is rarely a single-pass stick weld. A large structural or pipeline repair is a multipass deposit, each pass cleaned and ground before the next. In skilled hands this approaches dry welding quality, a substantial change from the rough single-pass welds of early commercial work, thanks to better consumables, tighter surface preparation, and inspection that confirms what is inside the joint.

For the shop-floor comparison of stick welding with other land processes, the breakdown in why SMAW still wins is worth a look. The reasoning underwater is largely the same, only the constraints are heavier.

Electrode Coatings and the Fight Against Hydrogen

Electrode selection is where most underwater metallurgy is won or lost, and the classic problem is hydrogen. A wet electrode must carry enough iron oxide in its coating to build the protective slag and gas bubble envelope that keeps water off the molten pool, and that iron oxide gives the wet weld higher diffusible hydrogen than a comparable dry weld. Hydrogen plus a brittle microstructure plus high residual stress spells hydrogen-induced cracking, the single biggest reason wet welds carry a risky reputation.

The industry answer has been the oxidizing, or rutile-type, coating, formulated with a relatively high proportion of iron oxide alongside other oxides. Formulation work has progressively reduced diffusible hydrogen and helped control porosity. The chemistry also affects slag behavior, and underwater slag floats upward, obscuring the puddle, the reinforcement, and the toe of the weld. Fluid slag lets the welder see and control the bead.

Consumable / process	Key underwater characteristic	Best suited for
Rutile (oxidizing) SMAW electrode	Low diffusible hydrogen, fluid slag	General wet repair, multipass work
Cellulosic SMAW electrode	High hydrogen and moisture pickup	Avoid wet; dry or hyperbaric use
FCAW (tubular wire)	High deposition, good mechanical properties	Long repetitive seams, mechanized feed
Bulkheads / clamp-type SMAW	Rigid, self-supporting	Vertical and overhead emergency joints

Outside the dive environment, the all-position rod guide covers consumable choices, and matching filler metals to base material is covered in alloy steel welding electrode selection and safety.

Polarity and the Role of the Dive Team

Polarity is familiar on land, but underwater it interacts with everything else. Generally a DC electrode positive (DCEP, reverse polarity) setting is used with many consumables because it gives deeper penetration and higher deposition rates, putting more heat into the work rather than the electrode. The choice follows electrode classification, joint configuration, and coating behavior in water, and must be fixed in the qualified procedure.

What differs underwater is not so much the switch setting as who controls it. The diver works in the dark, often in current, with poor visibility and an adrenaline surge. Current through the electrode holder, the work lead, and the water surface can put the diver into the circuit path in ways never seen in a shop, which is why grounding and the diver’s position relative to the return path are treated as a dedicated discipline.

FCAW and the Push Toward Mechanization

Flux-cored arc welding is a promising wet alternative because it delivers substantially higher deposition rates than stick, and because continuous wire fed from a surface-mounted cabinet removes much of the manual rod handling that slows a diver down. FCAW deposits also offer good toughness and strength for structural and pipeline work.

The catch is parametrization. An FCAW wire has a built-in flux, so shielding, slag formation, and heat input are coupled together. Running it underwater means holding a stable arc through a bubble envelope partly generated by the wire itself, in currents that move the water. The operating window is narrower than for stick, and nailing it on the first attempt is unrealistic, which is exactly the problem for a code-driven qualification regime. Research continues, but the qualification burden is why stick remains the default.

Flux-cored wire on land is covered in what actually works with aluminum flux core welding, a different material family with the same wire feed and shielding physics.

Special Underwater Equipment Considerations

Everything above depends on hardware that has to survive salt water, darkness, and pressure.

• Electrode holders and stinger leads. Wet-rated units with sealed bearings and corrosion-resistant housings; replace leads on schedule as conductor corrosion raises resistance.
• Welding machines. Constant-current sources; wet units add high open-circuit voltage for arc restarts and hot-start. The general welding equipment guide covers shop basics.
• Grounding and surfaces. The work lead clamps to the structure so current returns along the steel, not through the diver or tender.
• Slag, spatter, and visibility. Slag rises and clings to overhead surfaces, so shielding, drip plates, and debris traps are routine; a good camera and an aimed light are the cheapest quality tools.
• Chamber gas control. Hyperbaric work needs atmosphere control, monitoring, and a fire watch, covered in the codes section of this guide.
• Consumable storage. Electrodes and wire are moisture-sensitive; sealed storage and a dry room on the tender are not optional, since damp consumables cause hidden porosity.

Filler metal storage parallels how steel welding works, and surfacing and cleanup is close to air arc welding and carbon arc gouging practice. Process choice is less about the ideal arc than about what a team can qualify, control, and inspect.

Metallurgy, Defects, and Quality Control

Ask a welder what changes underwater and the answer is always the same: it gets cold, and it gets cold fast. Water is hundreds of times denser than air and pulls heat from the molten pool with a greed no shop floor can match. That explains nearly every metallurgical headache that follows, from hard, brittle heat-affected zones to cracking that appears days after the diver surfaces. It is why AWS D3.6M does not borrow the rules of ordinary structural welding.

Heat Loss and the Microstructure

A dry bead cools over tens of seconds, giving carbon time to diffuse out of the hardened zone. Underwater, the weld solidifies in a fraction of that time. The faster quench leaves less opportunity for tempering, so the heat-affected zone (HAZ) can retain high hardness, especially in steels with a high carbon equivalent. The result is a bead that looks sound but carries a brittle microstructure that may fail under service load. AWS D3.6M treats HAZ hardness as an acceptance criterion, which is why marine welding standards lean on hardness and mechanical qualification rather than appearance.

Hydrogen: The Number One Concern

Inside the arc, water molecules break apart. The weld atmosphere in a wet weld is rich in hydrogen, and hydrogen atoms are small enough to slip into the molten pool and the HAZ ahead of the solidification front. Combined with available carbon and tensile stress, they form hard, untempered martensite and, in service, hydrogen-induced cracking: underbead cracking that splits a sound-looking joint from the inside out. Reviews of wet welding metallurgy name hydrogen the single most restrictive factor on how high a carbon equivalent steel can be welded underwater.

Consumable design is the first line of defence, and it is why the electrode used at depth is not simply a shop rod in a wet suit. Flux chemistry and the slag-metal interface control how much hydrogen crosses into the weld, and AWS D3.6M restricts which electrode classifications may be used for wet welding. Underwater consumables use water-resistant coatings formulated for low-hydrogen performance. The same coating logic drives the selection and storage rules in our guide to alloy steel welding electrodes, and a damp container or wet glove can undo an otherwise sound procedure in seconds.

The Defects You Will Actually Find

Defect	Why it shows up underwater	Primary control
Porosity	A wet arc feeds nitrogen, oxygen and moisture into a pool that freezes instantly, trapping bubbles.	Controlled arc length, steady travel, dry consumables, chamber humidity
Hydrogen-induced cracking	Hydrogen diffuses into a fast-cooled, high-hardness HAZ ahead of the advancing pool.	Low-hydrogen consumables, restricted carbon equivalent, post-weld tempering
Lack of fusion / cold lap	High travel speed and a small mobile pool let filler metal roll ahead of the puddle.	Slow travel, correct rod angle, stringer beads
Incomplete penetration	Restricted access and electrode angle limit the root pass in a quenched joint.	Qualified joint design, proper prep, multipass planning
Undercut, slag inclusion, irregular profile	Arc turbulence destabilizes the pool and drowns the slag.	Short arc, steady travel, deliberate slag management

Mitigation: Heat Input, Pass Sequence, Tempering

• Control the heat input. Slower travel, a short arc and a tight, consistent bead give the pool time to wet out and the HAZ time to cool gently. Multi-pass stringer welds are preferred over wide weaving.
• Grind between passes. Every pass leaves a hardened layer. Back gouging or grinding it away exposes fresh base metal for the next pass and stops hardness stacking up. The mechanics are covered in our look at air arc welding and carbon arc gouging.
• Temper in place with induction heating. The most promising mitigation in the field. Research has shown induction heating on wet welds can temper high-hardness regions effectively, allowing qualification on steels with higher carbon equivalents that would otherwise be rejected on hardness grounds. The technique and its limits are documented in the technical literature on induction heating for underwater welding.

Quality Control in the Field

Inspection underwater is its own discipline. Visual examination comes first, followed by magnetic particle testing of fillet welds, and ultrasonic or radiographic methods where the code requires them. Hardness surveys on the HAZ are routine on higher carbon equivalent steels, and the paperwork matters as much as the probe results: weld maps, welder identification, electrode lot numbers and machine settings belong in the dive log so a repair can be traced years later. Recent technical reviews outline the promise and limits of these methods, including the push toward automated inspection.

Dry and hyperbaric welding solve most of these metallurgical problems at the source: no water to quench the bead, no wet arc atmosphere to load it with hydrogen. Wet welding remains the practical choice when access, cost and schedule rule out a chamber, and the rules above are what keep it defensible.

Codes, Standards, and Qualification

Underwater welding has a quality-control problem no land job has: the joint usually cannot be inspected, reworked, or repaired without enormous cost. On shore you grind out a porosity cluster and run another pass. Twenty metres down that means floating a chamber, repositioning a diver and losing a tide window. Codes turn that into a system: a qualified procedure, a qualified welder, defined acceptance criteria, and a record of what was done.

AWS D3.6M: The U.S. Baseline

AWS D3.6M:2017, the Standard for Underwater Welding, is the reference most U.S. underwater work follows. It is maintained by the AWS Committee on Welding in Marine Construction (D3) and covers wet welding in open water and dry welding inside a sealed chamber, governing welder performance qualification, procedure qualification, and acceptance criteria for the finished weld.

At its core are three quality classes, not one “good weld” grade but a ladder of acceptance limits tied to how much the joint matters.

Class	Typical application	Acceptance focus
Class A	Critical structural or fatigue-loaded joints	Tightest limits on porosity, undercut and cracks; NDE and witness points usually mandatory
Class B	Secondary structural connections	Moderate defect limits, restricted repair, NDE on a sampling basis
Class O	Miscellaneous or temporary joints outside the structural classes	Loosest limits, but a documented procedure and qualified welder are still required

Class selection is a design decision, normally fixed by the owner or design authority. It dictates how much repair is allowed, what non-destructive testing is mandatory, and which work must be witnessed.

Two details catch newcomers out. Depth and water temperature are essential variables, so a procedure qualified in shallow water does not automatically cover deep work. D3.6M is also strict on repair: a repair weld must meet the class of the original joint, and a repair of a repair carries tighter limits and often needs engineering approval first.

ISO 15614-9: Hyperbaric Wet Welding

Internationally, the framework comes from the ISO 15614 series. ISO 15614-9:2025 is the underwater part: it specifies procedure qualification tests for hyperbaric wet welding of steels in material groups 1, 2, 3 and 8, and defines weld quality levels A, B and Z.

Hyperbaric wet welding is the middle path between open water and a dry habitat. The joint sits in a pressurized, gas-filled bell while still surrounded by water, so the diver welds in a controlled atmosphere without the cost of a full dry-diving habitat. Level A is the strictest, B intermediate, and Z covers joints where a lesser quality is acceptable.

One clarification: ISO 15614-9 qualifies procedures, not people. Welder qualification runs through the ISO 9606 series and the relevant classification rules.

Classification Societies and Project Specifications

ABS, DNV, Lloyd’s Register, Bureau Veritas and TÜV each publish their own rules for underwater welding and repair. Many reference D3.6M or the ISO standards but add project-specific conditions on top.

Two patterns repeat. Class rules set thresholds for where wet welding is permissible at all, and scrutiny tightens as a joint approaches primary, load-bearing or fatigue-critical service. Many permit wet welding only on secondary or miscellaneous connections, prohibit it on primary structural members, and require the surveyor to verify the procedure independently before production welding starts.

Pressure-retaining work adds another layer: ASME BPVC Section IX, API 1104 for pipelines, and API RP 2 for offshore structures. Underwater pipeline repairs typically run to a qualified wet procedure referencing API 1104 plus class rules, documented and witnessed.

Ship repair adds flag-state and port-state hot work rules on top of class survey. Maritime and Coastguard Agency guidance on cutting and welding operations is a useful reference, and our guide to AJ Marine welding covers marine standards and safety practice more broadly.

What “Code Quality” Actually Means Underwater

Code quality is not a compliment about appearance. It is a checklist:

• A qualified procedure in range. The WPS covers the joint, position, depth band and consumable actually used. Change an essential variable and you requalify.
• A welder qualified for that class. Underwater joints are nearly all vertical, overhead or horizontal, so performance qualification is effectively all-position testing. Our all-position welding rod guide covers how electrodes behave out of position.
• Controlled consumables. Wet electrodes stay in sealed containers and are conditioned before use so moisture does not feed hydrogen into the weld.
• Acceptance testing by qualified personnel, read against the class limits.
• Documented repairs. Each repair is treated as a new weld with its own inspection and, often, its own witness point.

Outside that system you have a non-code repair, which is not automatically bad. Salvage crews use non-code welds to stop a leak and get a vessel to port for a permanent fix in dry dock. But it must be declared, and the owner has to accept it in writing.

The practical takeaway is that code-quality underwater welding is a discipline of records as much as of arc control. The codes tell you exactly what evidence to keep, and that evidence is the first thing a surveyor, insurer or regulator asks for. The science underneath keeps sharpening too, through the technical literature indexed in DOE’s OSTI database and recent peer-reviewed work in Welding in the World.

Applications, Benefits, and Limitations

Underwater welding is rarely a first choice in a project plan. It shows up because the alternative is worse: a platform shut in, a ship pulled into dry dock, a pipeline waiting for a weather window.

Offshore Platforms and Pipelines

Most underwater welding sits on the seabed and on offshore structures. Riser clamps, conductor casings, tubular joints, and pipeline spools are cracked by corrosion, anchor drag, dropped objects, and vessel traffic. Repairing in place keeps the asset producing: when a floating production unit is down, revenue loss runs into six figures a day, and lifting a damaged pipeline onto a repair barge can cost more than a patch welded in place.

Three approaches coexist: wet welds for temporary or non-critical repairs, clamped pipe repair (a sleeve or half-shell welded over the damage), and hyperbaric dry welds where the joint must meet code. Hydrocarbon service lines are the most demanding because the weld must be made around flammable gas, pushing work toward purged, sealed, or hyperbaric environments. The U.S. Department of Energy’s indexed collection of underwater welding studies documents this field.

Ship Repair and Ship Husbandry

Ship repair is where underwater welding earned its reputation. Hull cracks, punctures, propeller and rudder damage, shaft log seal housings, thruster nozzles, tank bulkhead failures, and plate renewal all happen below the waterline, on vessels that cannot always afford a graving dock. A wet repair alongside the quay can turn a multi-week docking outage into one shift.

The U.S. Navy’s manual on underwater cutting and welding covers this operational reality, and repair yards have written practice around it. For hull and structural repair standards, see AJ Marine Welding’s guide to marine processes and standards. Small aluminum craft are a special case: skiffs and workboats are common targets, though aluminum’s heat sensitivity and oxide film make the process far less forgiving than steel, as detailed in the guide to aluminum boat welding and strong hull joints.

Ports, Harbors, and Marine Terminals

Harbor work is a quieter but steady market. Divers repair sheet pile walls, quay ladders, fender panels, lock gates, bridge piers, dolphins, and mooring structures, plus submerged utility and outfall lines. These low-stress joints suit wet welding and keep costs down.

Nuclear and Decommissioning Work

Underwater welding appears in nuclear plant repair and decommissioning of retired reactors and fuel pools. Reactor pressure vessel shells, containment liners, and spent fuel storage basins have been repaired, cut, and re-welded in place, avoiding the cost of defueling and dismantling. It runs under its own regulatory regime, with welder, procedure, and inspection qualification far beyond ordinary fabrication.

Salvage and Emergency Response

Salvage is the impulsive end of the trade. A sunken vessel, a downed aircraft, or a collapsed bridge has to be cut and moved, sometimes in poor visibility and against a tide. Salvage welders cut and gouge as comfortably as they fillet weld, and underwater air arc gouging remains a familiar tool for plate removal.

The Business Case

The core benefit is financial. Dry docking a ship costs roughly what many small companies make in a year and adds weeks of schedule; an offshore repair that keeps a platform producing avoids both towing and downtime. Underwater welding lets owners repair in situ, extend asset life, and defer a dry-docking bill, which keeps the method economically attractive.

Where the Limits Bite

Depth is the first wall. Wet welding depends on the diver seeing the joint and on the arc behaving under ambient pressure, which limits work to shallow water, on the order of tens of metres. Visibility can be near zero in rivers, harbors, or North Sea sites, so lighting becomes part of the procedure.

Access is the second. The torch must reach the joint, the plate must be ground flat, and a clamp must fit around a pipe or member. Damage that closes a gap or distorts a plate often cannot be welded until it is cut out, and that removal is often the bigger job.

Time pressure is the third. Dive time, decompression stops, cold, current, and fatigue compress the working window, and the diver welds tired, often near the end of a bottom time. SMAW’s simplicity and forgiving arc are why it dominates the wet process.

Finally, not every joint may legally be wet welded. Repair limits come from the code: AWS D3.6M governs qualification of underwater welding procedures, and ISO 15614-9 sets procedure qualification testing for hyperbaric wet welding. Classification societies and insurers often require non-destructive examination and may still order dry repairs on critical joints. That is the honest trade: underwater welding is faster and cheaper, and sometimes the only option, but not always the qualifying one.

Safety, Hazards, and Best Practices

Underwater welding is among the most hazardous trades in metalworking. The danger sits in the combination: murky water, an arc that behaves differently underwater, electricity in a conductive environment, and a clock that never stops.

Before any operation, competent personnel complete a formal hazard review, check equipment, and confirm the plan covers diver exposure limits. The U.S. Navy Underwater Cutting and Welding Manual lays out this discipline for pre-job preparation, equipment checks, and safe procedures.

Electrical Hazards in a Conductive Environment

Saltwater is an excellent conductor, and fresh water conducts too. Every milliamp leaking from a damaged cable, wet connector, or electrode holder has a direct path back to the diver.

That fact drives equipment design. Underwater power sources are grounded, keep the electrode holder the only exposed conductor, and carry ground-fault and leakage protection that trips before a diver takes a dangerous shock. Many modern machines use an isolated or floating output.

• Inspect every cable and connector before the dive. Surface damage can hide a conductor failure that turns dangerous at depth.
• Keep splices and joints above water, made and dried on deck.
• Never use damaged or field-repaired electrode holders, the diver’s contact with the live circuit.
• Ground the work properly; a work clamp shortens the return path and limits stray voltage.
• Handle cables by hand; a snagged cable can drag a diver off position.

Diving and Decompression Risk

Underwater welding is almost always a diving operation, so the diver inherits every risk of working at depth: decompression sickness, nitrogen narcosis, oxygen toxicity, carbon dioxide buildup from exhaled gas, and the strain of cold, pressure, and heavy work.

Deco limits are not negotiable and not the welder’s to adjust. Schedules, ascent rates, and stops come from the dive plan and dive supervisor, so a half-finished weld gets stopped. Depth for a given weld is therefore an engineering decision.

Welding also produces fumes, and in some configurations hydrogen and other gases. The bubble shield over the arc can trap gas near the diver’s breathing zone, one reason shielding gas is deliberately limited underwater. A helmet, suit, and gloves for hot work are the baseline over a dry suit or hot water system.

Visibility, Time Pressure, and Human Factors

Visibility is often poor: sediment, darkness, and surface glare can cut working distance to a few inches. The diver works by feel, using the arc flash as momentary light, and the arc leaves eyes needing time to readjust.

With the dive clock running, the temptation is to rush, and that is where quality problems and accidents begin: rushed surface prep, skipped interpass cleaning, heat input pushed too far, or no inspection before the electrode runs out.

Human factors research names the culprits: task overload, fatigue, thermal stress, and the pressure of limited bottom time. Structured rest matters when a shift pairs a long surface interval with a demanding dive. Communication is another weak point, so the diver and surface tender share terms, hand signals, and a buddy check before the arc is struck.

Pre-Job and On-Dive Best Practices

The Navy manual and industry practice converge on a consistent checklist, and no weld starts until the hazard review is complete and everyone understands their role:

1. Write the dive plan: depth, duration, bottom time, gas, decompression schedule, procedure, abort criteria.
2. Complete a hazard review: confined space entry, entrapment, entanglement, visibility, surface traffic, structural hazards.
3. Inspect and function-test gear on deck: power source, ground-fault protection, cables, holders, comms, lighting, gas.
4. Set up to the governing code: under AWS D3.6M, wet welding carries strict depth limits the crew must know.
5. Control the surface condition. Underwater that means fighting marine growth and corrosion in poor visibility, and guidance on preparing aluminum for welding helps, since aluminum punishes contamination.
6. Manage consumables: keep electrodes dry, maintain holder contact, and use wet-process electrodes only if their classification allows.
7. Control heat input and interpass temperature; too much heat causes undercut, reinforcement, and cracking. Plan multi-pass work with grinding between passes.
8. Inspect before leaving the water. Visual and, where required, ultrasonic or radiographic examination must happen while access exists.

Training and Qualification

None of this works without competent people. Commercial divers welding underwater need a recognized underwater welding qualification plus the commercial diving certification their jurisdiction requires. Naval personnel follow their own training pipeline and the Navy manual, which stresses rigorous training, inspection, and hazard review.

Two failures recur in incident reviews: the belief that a competent surface welder can take a helmet and go underwater, and skipping inspection and documentation because the shift runs long. Underwater welding lives or dies on procedural discipline, because a failed repair underwater costs many times more than doing it right.

Respecting the conductive environment, the diver’s limits, and the bottom-time clock is what separates a safe, durable weld from an expensive problem waiting to surface.

Common Mistakes, Troubleshooting, and Practical Tips

Underwater welding punishes small decisions. Repairs get cut out and rewelded for avoidable reasons: a joint that was never truly dry, an electrode that sat in a wet container, a travel speed nobody controlled, an inspection nobody documented.

Preparation: where most repairs start and end

Surface preparation matters more underwater than in the shop. Scale, rust, oil, marine growth, and trapped seawater all increase porosity and drive hydrogen into the weld metal, where it combines with a hard, brittle heat-affected zone to produce delayed cracking. This is the most common root cause behind “mystery” underwater failures.

Good practice means abrading the joint to bright metal with a grinder or wire brush, blowing out crevices with air, and squeegeeing or wicking away free water. If a joint cannot be kept dry, the procedure needs a different plan, not a better hope. Nothing sitting in seawater should go straight into a live joint. Understanding why SMAW still dominates wet work helps here: electrode selection and storage are inseparable from the process.

Electrode handling and storage

Wet-welding electrodes are purpose-coated and moisture-sensitive. Damaged, pitted, or cracked coatings bubble and spit instead of shielding the arc, and electrodes stored in damp containers can absorb enough water to cause porosity. Store them dry, transfer them in sealed containers, and inspect every stick before use. Reject broken or crumbling coatings, and never return a wet-dragged stub to service. This all-position rod guide applies here too: a consumable is a process control, not a consumable part of the kit.

Heat input: the number nobody controls

Travel speed, electrode angle, arc length, and pass sequence determine how much heat lands in the joint. Too fast gives lack of fusion and a cold lap; too slow gives excess heat input, a wide hard HAZ, distortion, and greater hydrogen cracking susceptibility. Underwater wet welding cools fast, so the usable window is narrow and travel speed must be managed.

Watch the drag angle. Pulling the electrode forward lifts slag and improves wetting; pushing back invites undercut and poor bead shape. Keep the arc short and stable.

Interpass cleaning and multipass strategy

Grinding between passes is not optional. Slag, spatter, and the oxidized skin over the previous bead must come off before the next arc is struck, and in seawater those deposits regenerate fast. Brush the groove and confirm the surface is clean and as dry as possible before restarting.

Plan the sequence like a shop welder would: short, controlled passes with a back-step technique keep heat where you want it and limit distortion. Fill from the root outward, keep interpass temperature modest, and balance passes on both sides where possible. On thin or heavily restrained sections, light multipass passes separate a sound repair from a distortion nightmare.

Skipping inspection

Inspection is the first thing dropped when a job gets tight, and it is what ends careers. AWS D3.6M treats inspection as part of the process, and the ISO 15614-9 framework for underwater procedure and welder qualification makes testing a formal requirement. Document preparation, heat input, consumable batch, and inspection results while the diver is still in the water. If magnetic particle or ultrasonic examination is needed on a critical joint, plan it for a dry habitat or on the deck.

Hardness and in-situ tempering

Higher carbon-equivalent steels are the hardest case. Their welds develop hard, brittle microstructures and a crack-sensitive HAZ, and codes set maximum hardness values for this reason. Where the procedure cannot keep hardness within limits, in-situ tempering using local heating techniques such as induction heating may be needed, between passes or after the final pass. Plan it into the procedure, and the diver must see and control the temperature. Marine repair practice is covered in this guide to marine welding standards and safety.

Quick symptom reference

Symptom	Likely cause	Corrective action
Pinholes, porosity	Wet joint, damp electrode, unstable arc	Clean and dry, new electrodes, steady arc
Lack of fusion	Travel too fast, slag in groove	Slow down, grind and brush passes
Undercut	Wrong angle, excess speed, long arc	Correct drag angle, shorten arc
Heavy spatter	Long arc, wrong polarity, coated surface	Check polarity, short arc, strip joint
Hard, cracked HAZ	Excess heat input, high carbon-equivalent steel	Reduce heat, multipass plan, consider tempering
Distortion	Long beads, unbalanced sequence	Back-step, short passes, balance sequence

The strongest underwater welds come from crews that treat preparation, consumable control, heat input, interpass cleaning, and inspection as one disciplined system, not five separate concerns. Start with a qualified procedure from the AWS D3 committee on welding in marine construction, adapt it to the joint, and control what you can see.

Conclusion: The State and Future of Underwater Welding

Underwater welding is no longer a laboratory curiosity or a last-resort patch applied by a lone diver with a stick electrode. It is a code-defined discipline, practiced by trained diver-welders, governed by written standards, and used routinely on offshore energy assets, ship hulls, port structures, and submerged pipelines. That shift happened over decades of trial, failure, and research, and it is the single most important fact for anyone evaluating the craft today.

Where the industry stands right now

Through decades of research, wet welds can meet AWS D3.6M Class A requirements when the procedure is written, tested, and applied by skilled welders working alongside competent divers. That is the honest state of the art. It is not that every underwater weld passes, or that any competent welder can submerge a joint and repeat a shop result. It is that the joint can be brought to a defined, inspectable standard, with documented limits on depth, joint configuration, and structural criticality, and that inspectors and classification societies can rely on that documentation.

The framework keeps moving. AWS Committee D3 on Welding in Marine Construction reviews and revises the code as research and field experience accumulate, and hyperbaric wet welding has its own procedure qualification route under ISO 15614-9. Both matter because code compliance is what separates an accepted weld from a rejected one, and acceptance is what turns an emergency repair into a permanent asset.

The economics explain why wet welding dominates in practice. For shallow-water structural work, putting a diver and a welder in the water beats hauling a structure into dry dock by a wide margin. Where quality demands exceed what the surface can deliver, dry hyperbaric chambers remain the fallback, and guidance such as the cutting and welding reference material published for marine operators still reflects that preference for simplicity and speed over laboratory conditions. The dominant process remains SMAW, for reasons as much about handling and arc behavior in a conductive, cooling bath as about metallurgy. The strengths of the stick process are explained in our guide to why SMAW still wins.

The gaps that still hold the industry back

Three problems remain genuinely open, and they are why this trade still surprises people who assume it is a solved craft.

• Deeper water. As depth increases, the water column removes heat faster, raises hydrostatic pressure, and changes how the arc behaves and how bubbles form. Welds that are straightforward in a few meters of water become progressively harder to qualify and reproduce at greater depths.
• Hydrogen control. Moisture in the shielding environment introduces diffusible hydrogen into the weld. Rapid quench rates combined with hydrogen pickup drive hard, brittle microstructures and hydrogen-induced cracking, so electrode handling, drying, storage, and temper bead practice remain central. Selecting and storing electrodes properly is covered in our reference on alloy steel welding electrodes.
• Reliable non-destructive evaluation. Inspection is the least solved part of underwater work. Diver-held ultrasonic and magnetic particle inspection at depth, combined with radiographic work in chambers, is demanding and expensive, and diver eyesight and dexterity are finite. Better remote and vehicle-assisted inspection is an active research area, and federal technical literature such as the work catalogued by OSTI shows how long that effort has been running.

Where the technology is heading

The most promising direction for wet welding is flux cored arc welding. FCAW offers higher deposition rates, better arc stability in a moving conductive fluid, and a path to feeding wire from a surface or remote position instead of handling stick electrodes underwater. That combination makes it the leading candidate for productivity gains and, eventually, automation. Automation is not a new concept; underwater welding systems were already appearing in the technical literature in the mid-1980s, as shown in early underwater welding automation research, and the work has continued in more recent peer-reviewed reviews of the wet process.

Realistically, the near-term future looks less like a robot welding a pipeline and more like a better-supported human: ROV-mounted cameras and lighting, remote wire feed, improved dive ergonomics, and instruments that let a welder see and control the arc. Even so, the surrounding field matters. Proper setup of the equipment around the job, from gas supply through leads and ground clamps, is covered in our welding equipment guide.

The part automation will not solve

None of this progress changes the most important variable in underwater welding, which is the person holding the electrode. Diver-welders need formal welding qualification under the applicable code plus independent diving certification plus proven proficiency in a specific joint configuration, and employers have to keep that combination current. Surface preparation, electrode storage in a dry container, and heat input control decide the outcome far more often than process choice does. Our guides to all position rod selection and to marine welding practices and standards cover the shop and documentation side of that preparation, which matters just as much when the joint is underwater and aluminum hull repairs are among the jobs that cannot wait for dry dock.

The realistic summary is encouraging. Underwater welding is code-qualifiable in shallow water with proper controls, it is proven in the harshest commercial environments on earth, and the processes behind it keep improving. The remaining gaps, depth, hydrogen control, inspection, and automation, are engineering problems with funding and attention behind them. The part that will remain hardest to mechanize is disciplined workmanship, which is why training and standards will matter more in the next twenty years than they did in the last.

Frequently Asked Questions

What is the difference between wet and dry underwater welding?

Wet welding is done directly in water with no barrier between arc and water (usually SMAW/FCAW). Dry hyperbaric welding is done in a sealed chamber filled with gas at pressure, allowing welds closer to air quality but at much higher cost and with geometry limitations.

Is underwater welding as strong as welding above water?

With strict procedures, qualified consumables, and skilled welders, wet welds can meet code requirements for many structural repairs. However, without controls they are more prone to porosity, hydrogen cracking, and lower toughness than comparable air welds.

What codes govern underwater welding?

AWS D3.6M is the primary U.S. code covering underwater welding (wet and dry) with classes A, B, O. ISO 15614-9 covers procedure qualification for hyperbaric wet welding. Classification societies also have their own requirements.

What welding process is most common underwater?

Shielded Metal Arc Welding (SMAW) with waterproof electrodes is the most widely used process for wet underwater welding due to its simplicity, flexibility, and portability in complex underwater environments.

What are the biggest dangers in underwater welding?

Key hazards include electrical shock in conductive saltwater, hydrogen-induced cracking, rapid cooling leading to brittle welds, poor visibility, decompression risks for divers, and limited mobility/time underwater.

Can underwater welds be used for permanent repairs?

Yes, in many cases engineered permanent repairs are possible when performed under qualified procedures per codes like AWS D3.6M and accepted by the asset owner/classification society. The suitability depends on depth, steel, access, and criticality.

How deep can wet welding be done?

Wet welding has been demonstrated down to around 100 m in practice, with attempts deeper, but quality and hydrogen control become more challenging with depth. Most code-qualifiable structural wet welding is focused in shallower water (often under ~30 m) with strict controls.

What causes porosity in underwater welds?

Porosity is often caused by gas entrapment from water dissociation, inadequate cleaning, unstable arc, or improper technique. Proper flux chemistry, surface preparation, and controlled parameters help minimize it.

Do underwater welders need special training?

Yes. They must be qualified commercial divers and trained in underwater welding/cutting procedures, electrical safety, and often code-specific qualification. Organizations like commercial dive schools and Navy programs emphasize extensive practice.

Is flux-cored arc welding used underwater?

Flux-Cored Arc Welding (FCAW) is increasingly researched and applied in wet environments due to higher deposition rates and automation potential. It is less common than SMAW but represents the most promising direction for future development.