Pipeline Welding Description: Process, Codes & Inspection

At 6:40 on a cold morning in eastern Texas, a welder stands in a trench in front of a 24-inch pipe beveled and ready, a 3/32-inch stick electrode in one hand and a diesel generator humming behind him. Four hours from now the spread gets inspected, and within a few weeks a radiographer will shoot film of every girth weld he touched. That is the real nature of pipeline welding: not just an arc and a bead, but a chain of qualified procedures, controlled heat, impartial inspection and permanent records that decides whether the line can carry gas for the next forty years.

What Pipeline Welding Actually Is (and Why It Demands Its Own Rulebook)

Short answer: pipeline welding joins pipe, fittings and valves into a continuous, pressure-tight line that carries oil, gas, slurry or some combination of them under pressure for decades, through ground it does not control. That last clause separates it from everything else you might weld. A shop weld gets loaded once and inspected once. A pipeline weld takes a full-pressure hydrostatic test before it enters service, then sits there handling pressure, thermal movement and soil stress for the better part of forty years with nobody watching it.

Welds fall into two families. The overwhelming majority are girth welds, also called butt welds, joining pipe end to end around the full circumference. These are the joints a pipeline spread exists to make. The second family is fillet welds: where a flange joins the run, where a branch connection (weldolet or solet) taps into the main line, where a reinforcing sleeve is installed, or where a valve or fitting is attached. Fillet welds carry load differently and are judged on throat size, leg length and reinforcement rather than full penetration.

The consequence chain

Once the line is backfilled, it is buried. A slag inclusion, a root valley that was never fused, or a crack left in a weld is invisible for the life of the asset. It does not announce itself with a warning light. It stays there, slowly, until it becomes a leak into a wetland or a rupture in a neighborhood. By then the discovery is a headline, not a weld map entry.

So the industry builds a chain of controls around every girth weld: a qualified welding procedure, a welder performance-tested to that procedure, controlled heat input and preheat, documented consumable handling, scheduled nondestructive examination, written repair procedures, and permanent records tying it all together. This is a qualification-driven, inspection-backed craft, not an arc-welding free-for-all. Both the procedure and the welder are approved before a single production weld is made. If you have welded in a fab shop or your own garage, the pace will feel unfamiliar, because the paperwork is part of the job.

A sense of scale, using honest arithmetic

One hundred miles of 24-inch line laid at roughly 24-foot joint lengths works out to about 22,000 girth welds in a single project. The arithmetic beats a quoted number. Every one is a weld map entry with a welder identification number, a reference to the welding procedure specification used, and a nondestructive test report or a record of why one was not performed. A project that size is not a weld you finish. It is a production system that manufactures, tests and tracks welds continuously for months.

That scale is why the work is standardized down to the letter. In the United States, 49 CFR Part 192, Subpart E sets the legally enforceable minimum requirements for welding steel in gas pipelines, and 49 CFR Part 195, Subpart D does the same for hazardous liquid pipelines. One detail matters: Subpart E explicitly does not apply to welding performed during the manufacture of steel pipe or components. Mill welds fall under different standards than field girth welds, so the longitudinal seam in a pipe you receive at the yard sits on a different paperwork path from the circle you burn between two lengths.

Above the federal minimums sit the codes most operators work to: ASME B31.8 for gas transmission and distribution piping systems, and ASME B31.4 for pipeline transportation systems for liquid hydrocarbons, other liquids and slurries. ASME has been defining piping safety since 1922, and the B31 series is where the detailed requirements for materials, design, fabrication, assembly, erection, testing and inspection live.

What this article covers

Because the rulebook matters as much as the arc, this guide works outward from the metal: joint preparation and fit-up, process selection for a given diameter and wall thickness, what SMAW still does better than any newer process in the field, and why a modern girth weld is often a composite of two or more processes. From there it covers the qualification framework (procedure and welder, which beginners constantly confuse), heat input and preheat metallurgy, the inspection stack from visual through radiography and ultrasonic testing, the defect catalogue with causes and fixes, repair and in-service welding, and the safety and productivity reality of the job. New to the trade? Read the all-position rod and consumable guide alongside, since a girth weld can put you flat, vertical and overhead on one joint as the pipe rotates under the arc.

The Anatomy of a Pipeline Weld: Joint Preparation, Fit-Up and Groove Geometry

Before the arc is struck, the hardest work on a pipeline is already done. A rushed bevel, a dirty fusion face or a misaligned joint fights the welder on every pass and resurfaces as a rejectable indication in a radiograph.

Bevels, Groove Geometry and the Wall Thickness Rule

Most pipe ends arrive already bevelled, by a mill machine or one in the field. Bevel angle, root face and root gap come from the pipe wall thickness and the qualified procedure.

• Single-V groove: Standard for lighter wall. One bevel each side, welded inside out in root, fill and cap.
• Double-V groove: Heavier wall, where a single-V puts too much heat into one pass. Welding both sides halves the heat and tightens the heat-affected zone.
• Flare bevel or single-U: Wider included angle that spreads the pool and cuts filler metal, for thick sections with limited torch access.
• Root face: The unmelted land at the bottom of the bevel. It supports the molten root pool so it does not blow out, and it is a controlled dimension.
• Root gap: The space between pipe ends. Too tight and the root pass cannot be made cleanly; too wide and the weld metal bridges a gap it was never qualified for.

Cleaning the Fusion Faces

Weld surfaces must be free of dirt, grease, paint, scale, rust, oil and moisture, back about 25 mm either side of the joint. Neither shielding gas nor flux scavenges contaminants out of solidifying weld metal, so this is physical, not bureaucratic: an arc does not burn away mill scale or dissolve oil. It melts them into the puddle, where hydrogen, sulphur and hydrocarbons end up trapped as porosity or oxides.

Water in a stick electrode coating, a submerged arc flux or a wet shielding gas line is a leading cause of porosity and feeds hydrogen-assisted cracking later, so gas lines get leak checked, electrodes and wire stay in dry storage or holding ovens, and flux is re-dried when it takes on moisture. Stainless and aluminium add another layer: a heat-tinted oxide film or a trace of copper from a grinding wheel will seed a crack, so it pays to know how to prep aluminium properly.

Fit-Up, Alignment and the Dimensional Checks

Fit-up is the geometry check before welding: the ends are brought together, aligned, spaced and clamped.

• Root gap is gauged at several points around the circumference.
• Angular misalignment is the tilt between the two pipe axes; a high degree makes the welder chase a wandering root pass.
• Offset misalignment is the step where the ends do not share a centreline, putting the arc off joint centre and driving high wall thickness on one side.

Other checks cover wall thickness, ovality and straightness. Ovality matters because an out-of-round pipe presents a moving gap: the root gap at 12 o’clock is not the root gap at 6 o’clock. Misalignment is a common cause of rejectable ultrasonic and radiographic indications, since the defect was built into the geometry before the arc started.

Tack Welds: Made Right, Then Removed

Tacks hold the joint. They are made with a qualified procedure by a qualified welder, then removed and ground flush before the real weld is laid. That step gets skipped under schedule pressure. An unremoved tack is a stress riser and a crack initiation site that can stay buried under a cap pass for the life of the line.

Fillet Welds, Branch Connections and Sleeves

Flanges, branch connections and sleeves sit on the pressure boundary too. Fillet welds are measured by leg length and, more importantly, throat size, the effective depth that carries load. A long leg with a shallow throat is cosmetically impressive and structurally weak, and reinforcement, the extra metal above the surface, is limited because too much concentrates stress at the toe. Branch connections such as weldolets and solets follow their own rules, because a saddle branch meets the cylinder at an angle and that weld is not a simple fillet. Reinforcing sleeves are welded with the same discipline: their welds carry pressure loads for decades.

Arc Burns and the Federal Repair Rule

Arc burns happen when an electrode or tungsten touches the pipe outside the joint, leaving a hard, brittle spot in the parent metal. The rule is blunt. 49 CFR 195.226 requires each arc burn to be repaired by completely removing the notch by grinding, provided the grinding does not reduce the remaining wall thickness below the minimum required by the tolerances in the pipe manufacturing specification. If it does, a cylinder of pipe containing the entire notch must be removed and replaced. A ground surface may not be welded to the pipe or fitting, ruling out a cosmetic weld-over-the-burn.

Choosing the Process: GTAW, SMAW, SAW, FCAW and the Rest of the Toolbox

Choosing a welding process for a pipeline joint is not a matter of taste. Pick the wrong one for the wall thickness, the weather and your access, and you pay twice: once in the field, and again when the radiographic film comes back with rejectable indications.

Keep one framing in mind from the start. The process itself is never the deciding factor. What governs the job is the qualified Welding Procedure Specification (WPS) that matches the actual field condition, and the process is simply the vehicle that WPS authorizes. Federal rules require each procedure to be recorded in detail and followed in its entirety whenever it is used, so running a process outside that framework is not a shortcut, it is a violation (49 CFR Part 192, Subpart E).

GTAW: the root pass specialist

Gas tungsten arc welding (TIG) is the workhorse of the pipeline root pass. The arc and the filler metal are controlled independently, which lets the welder hold a very short arc, see the puddle and react to the joint as it closes. That control is why GTAW dominates stainless, duplex, nickel alloy and other high-alloy work, where cleanliness and heat discipline matter and where filler metal often costs more than the process itself.

Two conditions come with it. Backing purge is the first: many stainless and alloy roots need the backside swept with inert gas to keep oxygen off the molten metal, which means flow control, wind protection and an oxygen-deficiency hazard assessment. Tungsten risk is the second. Tungsten inclusions are associated with GTAW and plasma arc welding, and they come from dipping the tungsten into the weld pool or filler metal, letting it stick to the weld or base metal and then breaking it free, or running welding current above the tungsten’s rated capacity. A tungsten inclusion is a hard, unreactive speck inside otherwise sound metal, and it is an automatic reject.

SMAW: the field workhorse

Shielded metal arc welding, the stick process, grew out of a welding process developed in the early 20th century, and it remains the classic fill-and-cap choice on a pipeline girth weld. It needs no shielding gas, no flux handling and no sophisticated power source, which makes it the most portable process in the toolbox and a sound answer to a wet, windy, cramped right-of-way.

Its documented limitations are equally clear: low deposition rate, the need for a relatively highly skilled manual welder, relatively high defect rates compared with more automated processes, and high fume generation. That last point is a workplace issue as much as a quality issue, because stick produces far more fume per unit of deposit than a wire-fed process. If you are still building your electrode knowledge, the all-position welding rod guide explains how stick electrodes are classified, and alloy steel welding electrodes covers the low-hydrogen and Cr-Mo families that pipeline codes lean on in sour service.

SAW and the mill processes

Submerged arc welding produces very high deposition rates, which is why it dominates mill longitudinal seams, the long seam formed as pipe is manufactured, and why it shows up on large-diameter onshore pipe. The arc burns beneath a blanket of granular flux, so the operator cannot see it. No ultraviolet exposure, but no visual feedback either, and a well-documented set of SAW-specific surface defects including herringbone and pock marks. Flux handling and correct flux-burden height are critical, because an incorrect burden alters bead shape and creates the conditions for those defects (Kobelco’s defect reference explains the mechanism). In the field, SAW’s productivity gain is often offset by losing visual control of the pool.

At the mill, the choices widen. FCAW comes in gas-shielded (FCAW-G) and self-shielded (FCAW-S) forms and moves metal considerably faster than stick while working in fixed position. AWS publishes the consumable specifications behind all of this: A5.1 and A5.2 for carbon steel electrodes, A5.20 and A5.18 for flux-cored and GMAW wires, A5.36 for SAW fluxes, and A5.9 for GTAW consumables. GMAW/MIG is a mill staple for longitudinal seams. Plasma arc welding and laser welding appear at modern mills where automation and repeatability outperform operator skill. Electroslag welding (ESW) takes over at very large diameter and very thick wall, where a single pass deposits enormous tonnage of metal in one vertical position.

The composite girth weld

Most modern onshore girth welds are not one process. A representative 2026 study welded API 5L L415Q pipeline steel with a GTAW root pass and SMAW hot, fill and cap passes (Scientific Reports). That pairing is close to the industry default now: a clean, forgiving root from a process with full puddle control, then fast stick fill and cap layered on top.

What actually decides the process

Six drivers narrow the field: fixed position (a pipe does not move for you), diameter and wall thickness, weather and wind, access and productivity, and code acceptance. Gas-shielded processes, including GMAW, FCAW-G and GTAW, are wind-sensitive, and strong wind is listed as a direct cause of porosity because the shielding envelope is torn away and the arc atmosphere contaminates the pool. On an exposed right-of-way that means a windscreen, or a stop.

Process	Typical pipeline use	Key advantage	Key limitation
GTAW (TIG)	Root pass; stainless, duplex, high-alloy	Excellent puddle and heat control	Slow, costly filler, tungsten inclusion risk, purge needed
SMAW (stick)	Field fill and cap on girth welds	Portable, no gas or complex power source	Low deposition rate, high fume and skill demand
SAW	Mill longitudinal seams, large-diameter onshore	Very high deposition rates	No visual control; herringbone, pock marks, flux-burden issues
FCAW	Onshore girth welds, self-shielded where gas is impractical	Wire feed moves metal faster than stick	Wind sensitive when gas-shielded, consumables must stay dry
GMAW/MIG	Pipe mill longitudinal seams	Fast, automatable, repeatable	Wind sensitive, difficult away from the factory
PAW / Laser	Modern pipe mills	Automation and consistency	Capital cost, little field use
ESW	Very large diameter, very thick wall	Enormous single-pass deposit in vertical	Specialized equipment, strict position control

The Rulebook: API 1104, ASME B31.4, ASME B31.8, API 5L and 49 CFR

Almost every argument on a pipeline job traces back to a document, not opinion or shop folklore. The single source of confusion for newcomers is that “the code” is not one thing. It is a stack, and each layer answers a different question. Get the layers straight and the rest of the craft, from joint prep to radiography, starts to make sense.

Layer	Document	What it actually decides
Product	API Spec 5L	What pipe you are allowed to weld: grade, chemistry, mechanical properties, dimensions, manufacturing route, marking and traceability.
Design and construction	ASME B31.8 (gas), ASME B31.4 (liquid)	Design pressure, routing, materials, welding, preheat, heat treatment, examination and testing for the pipeline system as a whole.
Welding qualification and acceptance	API 1104, ASME BPVC Section IX	How a procedure and a welder become qualified, and how a finished weld is judged acceptable or rejected.
Enforcement	49 CFR Parts 192 and 195	The legally binding minimum in the United States, for gas (Part 192) and hazardous liquid (Part 195).

API 5L: the pipe, before anybody picks up a torch

API 5L is a product specification, not a welding standard. It defines the line pipe itself: grade, chemistry, mechanical properties, wall thickness, manufacturing route (seamless, longitudinally welded or spiral welded), mill seam inspection, marking and the material test reports that follow the pipe to the right-of-way. If the pipe is not API 5L, the code conversation shifts, because a line built from ASTM A106 or A333 pipe is typically governed by a different design code such as ASME B31.3, with acceptance criteria drawn from Section IX rather than API 1104.

API 1104: the welding and acceptance bible

The 21st edition of API 1104, Welding of Pipelines and Related Systems, is the document most pipeline welders will live inside for years. Its layout is worth knowing by heart, because the federal regulations cite it section by section. Section 5 covers procedure qualification: what a Welding Procedure Specification must specify and how a Procedure Qualification Record gets proved by test. Section 6 covers welder qualification, including the all-position test a pipeline welder must pass and the consumable classification that has to match the procedure, covered in our guide to all position welding rods and to alloy steel welding electrodes. Section 9 sets acceptance criteria and how results are interpreted. Section 12 governs in-service welding, the only lawful route to repairing a live line. Appendices A and B back these up, and 49 CFR 192.225 names them directly as acceptable bases for qualified welding.

The 21st edition also adds tables setting flash weld procedure qualification test specimen locations by pipe diameter, so the qualification coupon comes from a defined circumferential position on larger pipe rather than wherever it is convenient. Small detail, big consequence: it stops a process from being proved on a narrow coupon and then applied to a 36-inch girth weld.

ASME B31.8 and B31.4: design, construction and the welding chapters

ASME B31.8 governs gas transmission and distribution piping, including compressor stations, metering and regulation stations, gas mains and service lines up to the outlet of the customer’s meter set assembly. Its welding chapter runs in a logical order: 822 Preparation for Welding, 823 Qualification of Procedures and Welders, 824 Preheating, 825 Stress Relieving, and 826 Weld Examination Requirements. The B31.8-2022 revision, itself a revision of B31.8-2020, added a section on training and qualification of operating company personnel, a new section on developing a damage prevention program, and rewrote the plastic piping design formula section in its entirety.

ASME B31.4 covers liquid hydrocarbons, slurries and liquid anhydrous ammonia, reaching into corners the gas code does not: CO₂ piping sits in Chapter X, offshore requirements in Chapter IX, and slurries in Chapter XI. The 2019 edition reworked Chapter IX so its numbering matches the rest of the standard, which sounds cosmetic until you are cross-referencing a clause across two books at midnight.

49 CFR Parts 192 and 195: where the rulebook becomes law

Everything above stays advisory until a regulator writes it down. 49 CFR 192.225 requires that welding be performed by a qualified welder or welding operator under procedures qualified per Section 5 (except Note 2 in paragraph 5.4.2.2), Section 12, Appendix A or Appendix B of API 1104, or Section IX of the ASME Boiler and Pressure Vessel Code, and it states bluntly that the quality of procedure-qualification test welds must be determined by destructive testing. No coupon, no qualifier, no weld. 49 CFR 195.214 mirrors that for hazardous liquid lines and adds a paperwork duty that catches inexperienced teams: each welding procedure must be recorded in detail, including the results of the qualifying tests, retained, and followed whenever the procedure is used.

Finally, the acceptance rules tie the stack together. Under 49 CFR 192.241(c) and 195.228, weld acceptability is determined under Section 9 of API 1104, and Appendix A of API 1104 may not be used to accept cracks. That single line is why a radiographer can reject a girth weld for an indication that would otherwise be waved through as a fracture-mechanics special case. The rulebook is not paperwork around the craft. It is the craft.

Qualification: How a Welding Procedure and a Welder Become Legal to Work

Ask a new pipeline welder what they need to be legal on the job and you will usually get half the answer. There is a paper showing the pipe is sound, and there is a test to pass. Those are two separate systems, and confusing them causes real problems in the field.

One track qualifies the procedure: the written instructions specifying what is welded, how, and which variables are held fixed. The other qualifies the welder: proof that a person can deposit sound welds with that procedure, in the positions and to the standard the code demands. Pass one and you have a valid method. Pass both and you have a weld that survives a hydrostatic test, an inspector, and a failure investigation decades later.

Track one: the WPS and the PQR behind it

Most operators work to the model in ASME Section IX and API 1104. The Welding Procedure Specification (WPS) covers base and filler metal, joint details, current, voltage, travel speed, shielding gas, preheat and interpass limits, and heat input range. The WPS proves nothing on its own. It rests on a Procedure Qualification Record (PQR): test welds, reports and measurements showing the procedure actually produces conforming welds. API 1104’s 21st edition shows how tightly this is controlled, even down to coupon placement.

What the test welds prove is a metallurgical question, so the answer comes from destructive testing:

• Macroetch a cross section of the weld to reveal the fusion zone, check penetration and root, and look for cracking, lack of fusion or inclusions.
• Guided bend tests, face and root, to confirm ductility and that the weld and heat-affected zone survive the strain without cracking.
• Tensile testing, to confirm the weld metal is at least as strong as the base metal.
• Impact and hardness testing, wherever toughness or sour-service hardness limits apply.

Track two: qualifying the welder

Procedural competence is not personal competence. 49 CFR 192.227(a) requires each welder to be qualified under Section 6, Section 12, Appendix A or Appendix B of API 1104, or Section IX of the ASME Boiler and Pressure Vessel Code. A welder qualified under an earlier edition than the regulation lists may keep welding, but may not requalify under it. You can work under the book you learned; you just cannot renew your ticket with it.

A second path sits in 192.227(b): a welder may qualify to weld pipe operating below 20 percent of SMYS hoop stress by performing an acceptable test weld under Section I of API 1104 Appendix C. Separately, every welder making a welded service line connection to a main must first perform an acceptable test weld under Section II of Appendix C. Worth knowing, because qualification tests are all-position by nature, exactly where sound all position welding rod selection and disciplined electrode handling start to show.

Continuity of practice: the six-month rule

49 CFR 195.222(b) is the rule most people miss: no welder may weld with a welding process unless, within the preceding six calendar months, that welder engaged in welding with that process and had one weld tested and found acceptable under Section 9 of API 1104. Part 192 carries the same six-calendar-month logic. Continuity is tied to the process actually used, so a move from SMAW to FCAW can mean requalifying even though both look like “stick work” outside the trade.

What breaks continuity is time, a change of process, and stretches of work that never reach a tested weld. Requalification is not punishment; it keeps a welder’s hands honest on a job where the only defect that matters is the one nobody finds for ten years. If a welder is close to lapsing, schedule the requalification test before the gap rather than after it, and keep a weld log by date and process so nobody is guessing in the third week of a slow winter. Employers can track expiration like material test certificates.

Daily sampling and the permanent record

Qualification is sustained by evidence, and the pipeline record produces it daily. Under 49 CFR 195.234(d), at least 10 percent of the girth welds made by each welder during each welding day must be nondestructively tested over the entire circumference of the weld. Part 192 requires the operator to retain, for the life of the pipeline, a record showing the number of girth welds made, the number nondestructively tested, the number rejected and the disposition of the rejects. That is the audit trail a regulator, an insurer or an investigator will ask for decades later.

Heat Input, Preheat and Interpass Temperature: The Metallurgy Behind the Bead

A pipeline weld can look perfect and still be a liability. It passes visual inspection, sounds clean on film, and sets up the condition that produces a crack three years later at full operating pressure. A weld is a thermal event, and the heat put into the joint leaves a physical record in the steel’s microstructure. Heat input, preheat and interpass temperature decide whether the heat-affected zone (HAZ) keeps a fine grain structure and whether dissolved hydrogen gets trapped while the weld cools.

What Heat Input Actually Means

Heat input is how much energy the arc dumps into the metal per millimetre of weld, in other words the rate at which you are heating the steel.

Heat input (kJ/mm) = (voltage × amperage × 60) ÷ (travel speed in mm/min × 1000)

Arc energy is the same quantity expressed as kilojoules per unit length of weld, and most welding procedure specifications (WPS) use the terms interchangeably. Running 24 volts at 160 amps down a joint travelling at 150 mm/min gives roughly 1.54 kJ/mm. Change the travel speed and the number moves; double the amps and it doubles. Slow travel, wide beads and fat electrodes all push heat input up. Heat input is not amperage: a high current on a fast-traveling joint can still land inside the qualified range, while a modest current on a slow, hesitation-filled root pass blows straight through the maximum.

What the WPS Actually Controls

Every pipeline WPS addresses at least three thermal parameters:

• Preheat temperature — the minimum temperature the base metal must reach in the joint area before the arc is struck.
• Maximum interpass temperature — the highest temperature the weld metal or adjacent base metal may reach between passes.
• Heat input range — the minimum and maximum kJ/mm the procedure permits.

Preheat is not left to the discretion of the welder at the pit. ASME B31.8 and the corresponding provisions in ASME B31.4 govern preheating, so preheat values are written into the procedure and recorded on the weld map. Preheat does two jobs. It slows the cooling rate of the weld metal, giving hydrogen time to escape rather than being trapped, and it cuts the temperature differential across the joint, lowering residual stress where cracking is most likely.

Interpass temperature works the other way. Too high and the HAZ is reheated, grains grow and toughness drops. Too low and hydrogen accumulates, because the weld cools faster than hydrogen can diffuse out. Excessive interpass temperature coarsens HAZ grains, erodes toughness and gives a crack fewer grain boundaries to arrest it.

Preheat is measured, not assumed. On a cold morning in Texas the joint area has to be brought up to temperature and held, and the torch, the wire and the gas all have to be dry. That is why choosing electrodes rated for all-position, low-hydrogen work matters so much in the field. A filler that is not rated for the position, or that sat out in the truck overnight, quietly changes the hydrogen level and the crack risk profile of the joint you are building.

Hydrogen, Hot Cracking and Coarse Grains

The most common cracking problem in medium- and high-carbon steels is hydrogen-assisted cold cracking, and it is the reason low hydrogen practice exists. That practice means three things together: keeping electrodes and the weldment dry, controlling the thermal cycle with preheat and minimum interpass temperatures, and specifying consumables with controlled moisture and diffusible hydrogen content.

Hydrogen cracking is nasty because it is delayed. The weld passes inspection, cools, sits for days, then a crack appears in the HAZ with no obvious cause. It is worst in thick, highly restrained sections where high hydrogen, high residual stress and a susceptible microstructure all line up. As the Kobelco guide to welding defects explains, hydrogen embrittlement produces delayed cracking in the heat-affected zone of higher carbon steels, and the risk scales with restraint and section thickness.

Not all cracks are hydrogen cracks, and confusing the two leads to bad fixes. Hot cracking happens while the weld is still solidifying, in grain boundaries of the weld metal, driven by segregation and low-melting constituents in high-sulphur steels. That is a cleanliness and composition problem. Cold cracking happens after solidification and cooling, usually hours later, driven by hydrogen plus stress plus a susceptible microstructure. Grind out a cold crack and reweld without fixing hydrogen control and you get another crack.

Push heat input high for the sake of speed and you trade a cracking risk for a toughness risk, which matters on a line that sees cyclic pressure.

What 2026 Research on L415Q Showed

A 2026 study in Scientific Reports examined this trade-off on API 5L L415Q pipeline steel welded with a GTAW root and SMAW hot, fill and cap passes. Taguchi optimization across current, voltage, travel speed, heat input and reinforcement found that both the low and the high heat input condition produced joints meeting API pipeline welding specifications. The difference showed up in properties, not acceptability: low heat input gave better penetration consistency and a more uniform hardness distribution, while high heat input gave higher impact energy at the fusion line and in the HAZs.

Industrial validation on 250 pipeline girth welds then showed consistent welding quality, acceptable geometry and no critical defects on radiographic testing. That is the practical argument for treating heat input as a controlled variable rather than an accident of how fast the welder happens to move.

Sour Service and Hardness Limits

For pipelines carrying hydrogen sulfide and other wet contaminants, the equation changes. Cracking-resistant limits and hardness control come from sour-service standards such as NACE MR0175 / ISO 15156, which the project specification pulls in. The WPS must state those limits explicitly, including any temper or bead tempering steps, so preheat and interpass discipline becomes a matter of the final hardness reading rather than a rule of thumb. 49 CFR Part 195 Subpart D and the comparable rules in 49 CFR Part 192 Subpart E sit on top of that framework.

All of it serves one purpose: making sure the microstructure you never inspect directly behaves the way the code assumes. Heat input is tracked because it controls the thermal cycle, and the values live in the WPS because a qualified procedure is what allows the weld to exist.

I’ll rewrite this section to fit the 600–900 word range while keeping the heading, facts, and links intact.

Inspecting the Weld: Visual, Radiography, Ultrasonic, Magnetic Particle and Liquid Penetrant

Every pipeline weld gets looked at, and most get looked at more than once. The welder checks the bead as it cools. Then a qualified inspector performs a formal visual examination. From there, a selected share of welds moves into nondestructive testing, where film, an ultrasonic probe or a magnet reveals what the surface cannot show. These layers stack, and a weld must clear every one that applies before it is buried.

Visual inspection: the baseline nobody skips

49 CFR 192.241(a) requires that visual inspection of welding be conducted by an individual qualified by appropriate training and experience to ensure the welding is performed in accordance with the welding procedure and the weld is acceptable under the section’s criteria. In practice, that means someone trained to read a weld, not just someone holding a flashlight.

Visual inspection is where most pipeline welds are accepted and where most problems surface first: undercut, excessive reinforcement, overlap, misalignment, arc burns, a bead that wandered off the centerline. It is cheap, immediate and needs no equipment beyond good eyes, a light and a gauge. Its weakness is equally clear. It shows only what is on the outside, and it depends entirely on the inspector’s judgement.

When NDT becomes mandatory

The federal rule for gas transmission lines is blunt. Under 49 CFR 192.241(b), welds on a pipeline to be operated at a pressure producing a hoop stress of 20 percent or more of SMYS must be nondestructively tested. There are two exceptions, and both are narrow. Visually inspected welds approved by a qualified welding inspector need not be NDT’d if the pipe has a nominal diameter of less than 6 inches, or if the hoop stress is below 40 percent of SMYS and the welds are so limited in number that testing is impracticable.

What counts as an acceptable test is set in 192.243(a) and 195.234(a): NDT may be performed by any process that will clearly indicate defects that may affect the integrity of the weld. Trepanning, where a sample is cut from the weld for destructive examination, is specifically excluded as a substitute for 192.243(a) testing.

And the process has to be written down. Sections 192.243(b) and 195.234(b) require that NDT be performed in accordance with written procedures, by persons trained and qualified in the established procedures and the equipment used, with written procedures established for interpreting each test result. An inspector who cannot produce the written procedure behind a film interpretation is not qualified to make it.

How much gets tested

Mandatory testing does not mean every weld is tested. It means a defined share is, chosen at random, and the welder cannot pick the good ones.

• 192.243(d) requires that a random selection of each day’s field butt welds be nondestructively tested over their entire circumference. At least 10 percent in Class 1 locations, except offshore, and at least 15 percent in Class 2 locations, with further requirements in Class 3 and Class 4 locations such as crossings of major or navigable rivers.
• 195.234(d) and (e) for hazardous liquid lines require at least 10 percent of the girth welds made by each welder on each welding day to be tested, and that all girth welds installed each day in specified sensitive locations be tested over the entire circumference. Sensitive locations include onshore areas where a loss of hazardous liquid could reasonably be expected to pollute water, offshore areas, railroad and public road rights-of-way, overhead road crossings and tunnels, and incorporated subdivisions. Up to 10 percent may be exempt where testing is impracticable.
• Tie-ins get 100 percent. Under 49 CFR 192.227(d)(4), at pipeline tie-ins, including tie-ins of replacement sections, 100 percent of the welds must be nondestructively tested.

The methods, side by side

Method	What it detects	Field reality
Visual (VT)	Surface condition, profile, alignment, reinforcement	Universal, fast, subjective, surface only
Radiography (RT)	Volumetric flaws: porosity, slag, lack of fusion, cracks	Film or digital, permanent record, needs a radiation exclusion zone, slower
Ultrasonic (UT)	Internal flaws and precise depth location	Fast and field-friendly, operator dependent, blind or “unreadable” zones near geometry changes
Magnetic particle (MT)	Surface and near-surface cracks in ferromagnetic material	Cheap and sensitive, but works only on ferromagnetic materials
Liquid penetrant (PT)	Fine surface-breaking cracks	Requires a clean, non-porous surface, and cannot detect subsurface flaws

Radiography remains the reference method at scale. A representative 2026 field-scale validation of 250 pipeline girth welds, published in Scientific Reports, evaluated quality by radiographic testing and reported consistent weld geometry with no critical defects, a good illustration of why RT is still the acceptance method many operators default to for girth welds.

Acceptance criteria and re-testing repairs

Finding an indication is only half the job. The other half is deciding whether it matters, and that decision is written down too. The API 1104 Section 9 acceptance criteria set limits for porosity, slag, undercut and other indications, while ASME B31.8 and B31.4 add the design and construction requirements for liquid and gas lines. Cracks are never acceptable in any of them.

Repairs do not get a pass because the original weld failed. Once a defect is gouged out and rewelded, the repaired segment must be inspected to ensure its acceptability under 49 CFR 192.245, and repair welds fall under the same nondestructive testing requirements as any other. Removing a reject usually means carbon arc gouging back to sound metal, then rebuilding with a qualified procedure.

The Defect Playbook: What Goes Wrong, Why, and How It Gets Fixed

A girth weld that gets buried, pressurized and forgotten for forty years has to be right the first time. When an indication shows up on radiograph or ultrasonic exam, the reaction is rarely “look closer.” It is “what caused this, and who signs the repair.” Pipeline weld defects are not random. Nearly all trace to a short list of causes: too much heat, not enough shielding, a dirty joint, a bad travel angle, a rushed consumable change.

The Kobelco welding defect reference and classroom materials on welding discontinuities and defects organize the subject the same way, by cause and by corrective action. That is the format that matters when a weld must be fixed before the trench closes.

Three Words to Keep Straight

An indication is a response from a nondestructive test that requires interpretation. A discontinuity is an intentional or unintentional interruption in the physical structure or configuration of a part. A defect is a discontinuity that fails the applicable acceptance criteria. Porosity inside the API 1104 acceptance limit is an indication, not a defect; the same porosity a hair beyond the limit is a defect and triggers the repair rules in 49 CFR 192.245. A radiograph shows only an indication. A qualified person decides what it means.

Undercut

Undercut is a groove melted into the base metal next to the weld toe. The causes are almost entirely parameter and geometry related: amperage too high, electrode manipulation or travel speed too fast, arc length too long or arc voltage too high, and inappropriate electrode travel and work angles. On submerged arc welding, wire tracking too close to the groove face adds to it. The fixes mirror the causes: reduce amperage, shorten the arc, correct the electrode angle, adjust travel speed and shift the wire tracking location. Undercut is a geometry defect, so it can fail acceptance even when nothing volumetric is present.

Porosity

Porosity is gas trapped in the weld metal. Documented causes: rust, oil, paint or moisture on the joint fusion faces; high sulfur content in the base metal; moisture in SMAW coatings, SAW fluxes or shielding gas; too little shielding gas or too low a flux-burden height; strong wind; and too much welding amperage, arc length or arc voltage. The fixes follow the same logic: clean the joint back to bright metal, redry consumables, control gas flow and shelter the arc from wind. Moisture is the recurring villain, which is one honest reason why SMAW still wins in the field holds up against gas-shielded processes.

Lack of Fusion and Slag Inclusions

Lack of fusion means the prepared root edge of the bevel was not consumed by the weld. It is generally tied to too-fast travel speed, inadequate heat input, wrong bevel angle or contaminated surfaces. Slag inclusions are trapped oxide and flux, and the contributors are improper welding technique, drag versus push technique, improper joint preparation, narrow deep grooves and unsuitable travel speed. Slag must be gouged and/or ground out and reweld with proper joint preparation and technique. Carbon arc gouging is the standard field tool for opening up a weld for repair.

Arc Strikes, Tungsten Inclusions, Cold Roll and Overlap

Arc strikes are attributed to welder error. Properly disengage the stick electrode from the electrode holder on SMAW and stop current to the torch before breaking the tungsten free on GTAW. A strike outside the joint leaves a hard, unmelted crater that acts as a stress riser, so it gets gouged out and repaired like any other defect. Tungsten inclusions come from a broken or dipped tungsten, an unstable arc or a dirty gas cup. Cold roll and overlap are caused by improper welding parameters, and welding documentation cautions against washing the weld face in most cases, since that invites contamination and erratic fusion.

Surface Irregularities and Cracking

Surface irregularities include uneven weld bead ripples, excessive reinforcement, uneven-leg fillet welds, undercut, overlap, herringbone on SAW, pock marks on SAW and mouse foot marks on SMAW. They read as cosmetic, but API 1104 sets limits on reinforcement and profile, so “cosmetic” can still be rejectable.

Cracking is the one nobody wants to find, and it splits into two families. Hot cracking happens during solidification and links to dirty surfaces, wide bead spacing and wrong preheat. Cold cracking happens later and is hydrogen-assisted, so it is worst in thick, restrained joints and in high-carbon or sour service steels. Control means dry consumables, the preheat and interpass limits written into the WPS, and for sour service, hardness limits drawn from NACE MR0175/ISO 15156.

Defect to Cause at a Glance

Defect	Usual cause	First corrective action
Undercut	High amperage, fast travel, long arc, poor work angle	Lower amperage, shorten arc, correct angle
Porosity	Contamination, moisture, wind, weak shielding	Clean joint, redry consumables, shelter arc
Lack of fusion	Fast travel, low heat input, wrong bevel	Slow down, raise heat input, re-prep groove
Slag inclusion	Poor technique, drag versus push, narrow groove	Gouge or grind out and reweld
Arc strike / tungsten	Contact outside the joint	Remove crater, repair as a defect
Cold roll / overlap	Improper parameters, wandering puddle	Retune parameters, steady travel
Surface irregularity	Unstable travel, heavy reinforcement	Grind to profile, verify limits
Cracking	Hydrogen, restraint, hot cracking conditions	Excavate, requalify repair, review WPS

A Troubleshooting Flow That Works on a Real Job

1. Read the indication correctly. Method, position on the circumference and orientation all matter. A tight indication at the toe behaves differently from one at the center of the weld.
2. Identify the process and position. The same-looking indication means different things on GTAW, SMAW and SAW.
3. Match it against the cause list for that process before blaming the welder.
4. Check consumables and environment. Wet electrodes, damp flux and a stiff crosswind explain a surprising share of rejects.
5. Decide the repair path from the acceptance criteria, not from how the bead looks.

I’ll rewrite this section to fit the 544–816 word range while keeping all facts and links intact.

Let me check the current word count and tighten it.Word count is now 808 (within 544–816), all 8 original links preserved, with tightened prose and added safe-target attributes on external sources.

Repairs, In-Service Welding and the Life of the Weld After the Backhoe Leaves

A girth weld does not stay new. It gets handled and tested, and sooner or later something shows up in a radiograph that nobody likes. What happens next is not a judgement call made on the spot. Federal rules decide who can cut, who can weld, what must be qualified, and what must be documented.

The Federal Repair Rules: 49 CFR 192.245

A weld that is unacceptable under the acceptance criteria has to go. 49 CFR 192.245(a) offers only two options, remove or repair, and draws a hard line on cracking. Except for welds on an offshore pipeline being installed from a pipeline vessel, a weld must be removed if it contains a crack more than 8 percent of the weld length. On a 20-inch circumference that is a tiny allowance, so one serious crack in a long weld usually takes the whole weld out.

It is the reason API 1104 Appendix A may not be used to accept cracks, even though Appendix A is the tougher acceptance criteria set. Crack repair falls under 192.245(c) instead.

Section 192.245(b) sets the standard for the work itself: remove the defect down to sound metal, preheat the segment if conditions exist that would adversely affect the quality of the repair, and inspect the repaired segment to confirm acceptability. Excavating a flaw down to clean parent metal is why carbon arc gouging and grinding are constant companions to the repair welder.

Finally, 192.245(c): repairing a crack, or repairing a defect in an area already repaired, must follow written repair procedures qualified under 192.225, and must ensure the minimum mechanical properties specified for the original welding procedure are met when the repair is complete. A repair weld cannot be a downgrade of the weld it replaces.

In-Service Welding: Repairing a Line Still Carrying Product

Taking a transmission line out of service to fix a weld costs days and triggers contractual headaches. Operators strongly prefer repairs that do not require taking the line out of service, a preference the ASME B31.8 transportation pipelines chapter states directly. That is why 49 CFR 192.225 and 49 CFR 195.214 both permit in-service welding procedures qualified under API 1104 Section 12 and Appendix B.

The tradeoff deserves honest accounting. Welding a live, pressurised, flowing line adds risk a cold line does not: pressure control, possible product release, restricted access, and a fire or toxic exposure event beside a hydrocarbon or H2S stream. In-service work needs a procedure qualified for those conditions, disciplined leak testing, and full NDT of every weld the repair creates.

Sleeve Repair Versus Full Cut-Out

Two paths dominate. A full cut-out removes the damaged section and replaces it with new pipe, the cleanest metallurgical answer but at a heavy price: excavation, two field girth welds and an in-service tie-in. A sleeve or reinforcing repair fits new material over the flawed weld and welds it in place, avoiding full excavation. It is faster and often far cheaper in a marsh or under a road crossing.

Approach	Strength	Trade-off
Full cut-out	Cleanest metallurgical result, new pipe throughout	Excavation, two field girth welds, longer outage
Sleeve or reinforcing repair	Avoids full excavation, fast, suited to crossings	Adds its own welds, each qualified and inspected

A sleeve is not a way around the rules. It carries its own welds, made by qualified procedures and inspected. The repair is only as good as the welds holding it in place.

Documentation, Traceability and the Long Tail

Every repair builds a paper trail: original weld number, defect, removal method, repair procedure, preheat record, NDT results, disposition. That file is what a regulator reads first and what an operator needs the next time the line comes apart. Girth welds are also a signature feature for in-line inspection tools, and a tool can only report meaningfully if the weld map, weld numbers and NDT records stay accurate. Those records remain in the operator’s file for the life of the asset.

Forty Years Later

A weld that passed radiography on day one still has to survive corrosion under insulation, sour service degradation, cyclic pressure and thermal movement, and the gouging damage of an excavator that comes close decades later. Operators run in-line inspection tools on a schedule, and girth welds appear as identifiable features the tool can size. A corrosion indication or crack found at a weld returns to the same framework: remove, repair, requalify, document. The backhoe leaves. The record stays.

Safety, Productivity and How to Build a Pipeline Welding Career

Safety on a pipeline spread is not a poster on the trailer wall. It is process control: keep the weld sound and the crew out of the hospital. The safest welders look busy, checking an oven log, shielding a joint with a tarp, reading a purge meter. None of it looks heroic, and it is why their welds pass radiography.

OSHA sets the floor, not the ceiling

Federal law on welding, cutting and brazing lives in 29 CFR 1910.252 and 1910.253, covering eye and face protection, fire watch duties and fume extraction, which matters most on stick welding. Pipelines stack more on top: OSHA trench and excavation rules, federal pipeline safety regulations, and the operator’s own safety management system. Where requirements differ, the stricter one governs.

Welding stainless steel, hard-facing or chrome-bearing consumables releases fume containing hexavalent chromium, a confirmed carcinogen. Local exhaust ventilation and properly selected respiratory protection are not extras on stainless or duplex work. They are the reason the job is allowed to happen.

Weather protection is a quality control

49 CFR 195.224 requires that welding be protected from weather conditions that would impair the quality of the completed weld. That one sentence ties safety to metallurgy. Wind breaks shielding gas and disrupts the arc in gas-shielded processes, a documented cause of porosity. Rain, condensation and damp electrodes introduce hydrogen, the ingredient behind hydrogen-assisted cold cracking in thick-section joints. A tent over the joint is a process parameter, not comfort.

The hazards that catch experienced people

• Trenches and excavations: cave-ins, spoil pile setbacks, utility strikes, daily competent-person inspections.
• Confined space: pit entry requires a permit, atmospheric testing and a trained attendant.
• Traffic and right-of-way: flaggers, tailboards, high-visibility PPE, controlled speed.
• Working at height: fall protection on slopes, scaffolds and pipe racks.
• Hot work: permits, fire watch and extinguishers on the truck.
• Nitrogen backing purge: an oxygen-deficient pit can kill quickly. Verify purge gas and monitor the pit.
• Radiography: exclusion zones, surveys and controlled access; ultrasonic sets bring high-voltage field hazards.
• Electrical: generators, leads and wet ground in an open trench.

Productivity is about rework, not speed

The expensive failure on a pipeline spread is not a slow bead. It is a rejected girth weld: excavation, gouging, rewelding, re-inspection and days of lost schedule. Qualified procedures, disciplined consumable handling and scheduled NDT attack that cost. Redrying electrodes and fluxes, controlled filler metal storage and heat lot tracking are traceability requirements, not housekeeping. When electrode and consumable selection goes wrong, the problem surfaces at the radiographer.

A 2026 study on API 5L L415Q pipeline steel models the improvement well. Researchers used Taguchi experimental design to optimize current, voltage, travel speed, heat input and reinforcement, then validated across 250 industrial-scale girth welds, with radiographic results showing acceptable geometry and no critical defects. Read the published findings. Parameters tuned by experiment beat parameters inherited by habit.

Shop equipment helps within limits. Holding ovens keep consumables dry; without them, redrying ovens and controlled storage become the bottleneck. Hot wire systems lift deposition rate on SMAW but complicate consumable handling and add fume near the arc. Preheat torches and positioners improve consistency; neither replaces a qualified procedure.

Wherever you are now, here is your next move

If you are a student:

• Learn the code stack in order: API 5L, ASME B31.8 or B31.4, API 1104, ASME Section IX, then 49 CFR 192 and 195.
• Get groove practice plate time before field work. A root pass is easier to learn on plate.
• Study career paths in other welding sectors to see where pipeline credentials travel.

If you are a new pipeline welder:

• Know your qualified positions and heat input range well enough to explain the preheat minimum.
• Protect consumables and record lot numbers from your first day.
• Track your continuity of practice. The federal six-month rule is real; losing it costs requalification.
• Ask to see the written repair procedure before you need it.

If you run QC or NDT:

• Show welders their own radiographs, not just a reject number.
• Track sampling compliance per welder per day, not per crew.
• Verify purge records, oven logs and lot traceability on every weld map.

The short version

Pipeline welding joins pipe and fittings into a pressure-tight line under procedures qualified in writing, inspected impartially and recorded for the life of the asset. GTAW roots, SMAW fill and cap, SAW and FCAW are only the visible layer. Underneath sit the codes, heat control, NDT and safety rules that keep the crew and the line in service. Get those right and the arc takes care of itself.

Frequently Asked Questions

What is pipeline welding, in simple terms?

Pipeline welding is the joining of pipe, fittings and valves into a continuous, pressure-tight line that carries oil, gas, water or slurry. Most of it is girth welding — running a qualified weld all the way around the circumference where two pipe ends meet — plus fillet welds at flanges, branch connections and reinforcing sleeves. Unlike general fabrication work, every pipeline weld must be made under a qualified procedure by a qualified welder and is subject to inspection and permanent recordkeeping.

How is pipeline welding different from structural or shop welding?

Three things separate it. First, the consequences: a pipeline weld carries internal pressure, cyclic surge, thermal movement and soil movement for decades, and the line is proof-tested before it enters service. Second, the paperwork: procedure qualification, welder qualification and traceability are legally required, not optional. Third, the environment: welds are made outdoors, in trenches, in wind and rain, often out of position, on a schedule driven by a construction spread rather than by shop convenience.

What does API 1104 actually cover?

API 1104, ‘Welding of Pipelines and Related Systems’, is the pipeline welding standard. It contains the rules for qualifying welding procedures (Section 5), qualifying welders (Section 6), accepting and interpreting weld inspection results (Section 9), and welding in service (Section 12 and Appendix B), with Appendix A covering additional acceptance scenarios. U.S. federal regulation 49 CFR 192.225 and 195.214 point operators back to these sections or to ASME Section IX.

Which welding processes are used on pipelines?

GTAW (TIG) is common for the root pass, especially on stainless, duplex and high-alloy grades. SMAW (stick) is the traditional field process for fill and cap passes. SAW (submerged arc) dominates mill longitudinal seams and large-diameter onshore pipe, while FCAW and GMAW are used at pipe mills and for semi-automated field welding. Many modern girth welds are composites — a GTAW root with SMAW fill and cap — with the process chosen to suit diameter, wall, weather and position.

Does a pipeline welder need to be certified?

Qualified, rather than ‘certified’. Under 49 CFR 192.227 and 195.222, welders are qualified to API 1104 Section 6 or ASME Section IX, and qualification is tied to the specific process and conditions. It also expires in practice: a welder may not weld with a process unless within the preceding six calendar months they have welded with that process and had one weld tested and found acceptable under API 1104 Section 9.

What is the difference between radiography and ultrasonic testing for pipeline welds?

Radiography produces a permanent film or digital image, so it is an auditable record, but it is slower, needs radiation safety controls and an exclusion zone, and it is less sensitive to tight planar defects. Ultrasonic testing is fast, works well in the field and detects planar defects well, but it depends heavily on operator skill, gives no permanent record, and can leave ‘unreadable’ areas that require a supplementary method. Most operators use a combination, with acceptance set by API 1104 Section 9.

What is heat input and why does it matter so much?

Heat input is the energy delivered per unit length of weld, usually expressed in kJ/mm, and it is calculated from voltage, amperage and travel speed. Too much heat causes coarse grain in the heat-affected zone and can promote cracking; too little leaves poor penetration and lack of fusion. It also drives cooling rate, so the qualified heat input range in the WPS is what actually controls weld metal toughness and hardness.

Can you weld on a pipeline that is already in service?

Yes, but it is a regulated, high-consequence activity. In-service welding procedures are qualified under API 1104 Section 12 and Appendix B, and both 49 CFR 192.225 and 195.214 permit their use. The line must be depressurised or controlled to an agreed limit, the work needs written repair procedures qualified under 192.225, and the repair welds are fully inspected before the line returns to service.

What are the most common pipeline weld defects and their causes?

Porosity, undercut, lack of fusion and slag inclusion dominate. Porosity usually traces back to contamination or moisture on the joint faces, wet consumables or flux, insufficient shielding gas, wind, or excessive amperage and arc length. Undercut comes from excessive heat, too-fast travel speed, too long an arc or an incorrect work angle. Lack of fusion means the prepared bevel edge was not consumed by the weld, and slag inclusion is usually a technique and joint-preparation problem.

What should I study to work in pipeline welding?

Start with the structure of the work: girth welding and joint preparation, then the process fundamentals of SMAW and GTAW, then the rules — API 1104, ASME Section IX and the relevant B31 code — because the codes decide what a welder is allowed to do. Add nondestructive testing basics, metallurgy and heat control, and the safety side: OSHA requirements, fume control, purge gas handling and trench safety. Working alongside a qualified pipeline crew on a live project is worth more than any single certificate.