Many welding shops that pursue high-spec contracts assume the issue is human. The welder is responsible for ensuring work meets spec. If the welder isn’t good enough, hire a better one. If you’re still failing x-ray, have someone with more stamps look at it. Hopefully, the numbers will work out at the end of the job. That’s a difficult way to accidentally run a business, and it’s also 100% backwards. Quality, especially on demanding fabrication work, is an engineered result you should imagine in detail before anyone strikes an arc, not an attribute you cross your fingers and hope for in the booth.

The best fabrication shops in the world, the ones that pass third-party audits and win tier-one work long term, treat quality this way. Fill in this form and dozens of other variables that affect a weld’s mechanical properties will be documented, controlled, and tracked. The biggest cost of an inspection or test plan isn’t the labor to execute it, but the mistakes it finds that you make, the rework required to fix those mistakes, the waste they create, and the potential liability if that waste somehow ends up in a deliverable product.

 

 

Quality Starts Before the Quote is Signed

The costliest errors a welding business can make come long before production begins. These happen during the contract review and estimating phase, and although they’re not as visible as welding or cutting errors, they can cost you just as much, if not more.

If you don’t evaluate weldability and determine the critical weld classifications for the bid, followed by a review to establish which welding standards apply, you’re essentially guessing at the final costs because you’re working with the wrong information.

Fast-tracking contract review can lead you to either walk away from work you think you can’t do, or, worse, paint you into a corner to commit to work you can’t afford to certify. If you’re a small shop in a region where many jobs don’t require NDT, your procedures will have to be top of the range in all other respects to outbid competition that can self-certify under their existing procedures.

Reducing estimating time may or may not be the win you’re looking for. Reducing estimating cost certainly is, but only if you maintain estimating accuracy. Efficiently determining the contract review information you need to gather can cut estimating time without cutting corners that come back to bite you.

A Documented WPS Beats “Experienced” Welders Every Time

If you don’t have a WPS qualified by a PQR for the specific materials and conditions of service you’re welding under, you’re effectively just hoping the combination you’re using works. The more expensive or critical the application, the less likely anyone is to be willing to take that chance resulting in a rejected weld.

If the as-welded joint doesn’t meet required mechanical or NDE standards, you don’t get a do-over. At best, you’re looking at a costly, schedule-busting repair. At worst, it’s a write-off. That’s why any reputable welding operation starts every job by making sure they’re following a qualified WPS to the letter, before a single arc is struck.

Certification is a Tracking Problem, Not a Filing Cabinet Problem

Welder certification, such as under AS/NZS 3992, isn’t a box you check once and forget. It’s specific to the welding process, the position, and often the materials being joined, and must be regularly renewed for a welder to continue working on that process. If a welder hasn’t welded using a process on a material in a certain position for six months, they are no longer certified to make a welded joint using those specific parameters. If they were unqualified but still made a weld and something went wrong, you would not be covered by your insurance.

Although this kind of knowledge makes intangible assets swell in a business, many workshop supervisors can only guess at how many actual welds in a month each of their staff make commercially. Keep your customers and let them know, with pride (backed up by evidence), the level of skill of the team you put on their work.

Equipment Selection is a Quality Decision, Not a Budget Line

Here’s where a lot of quality-first thinking stops short. Many high-quality facilities meticulously document their welding procedures but then overlook the fact that their welders are using equipment not designed to meet the demands of those same procedures. These are effectively industrial athletes being handed running shoes suitable for quick trips between the couch and the fridge.

The job of the welder isn’t just to reproduce the sample in a procedure though, which is why not enough attention is paid to the device they use to create that weld. To put it simply, the quality and consistency of a weld depend on output power and arc stability, these aren’t just trivia. Does the power supply generate the output you need to make a certain weld according to the WPS? Does it change as the machine heats up and cools down mid-shift in a sawtooth pattern under sustained load? That will dictate how uniform your crystals end up between when the machine is cold versus when it’s been working for hours.

Similarly, if your welding process requires starting thick and automatically tapering down based on the feed your power supply isn’t giving you because it’s too hot, you have some serious problems. The welder is still going to adjust the power, positive or negative, based on whether they want to crank the output up or down, but that’s not a procedural variation, because the resulting weld quality is just a response to the power supply no longer being able to do what the welder needs it to.

All these continuous adjustments exact a price on your procedure because most welders will just do them and assume that’s their job. It may even be that you haven’t realized what compromises you’re making until you finally test for the difference. Shops running sustained critical production on high-spec contracts often standardize around industrial-grade platforms like kemppi welders specifically because dependable arc performance under continuous load reduces the number of times an operator has to improvise around a machine’s limitations. Document the variations or lack thereof in the WPS, and now you have the first procedural violation in front of the inspectors.

Build Inspection Into the Schedule, Not Onto the End of it

It’s too late to identify and correct issues after the task is done, as you’ll only be able to see how much rework needs to be done. A workflow that prioritizes quality establishes hold points throughout the production process, such as verifying fit-up before welding, conducting in-process visual checks at specific points, and planning NDT (radiographic, ultrasonic, MPI, or dye penetrant testing) to determine if work can proceed to the next phase.

Hold points are not red tape. They are designed to prevent a poor fit-up from turning into a welded joint that must be removed later. Each hold point must have a specific pass/fail criterion based on the relevant code, and someone must be authorized to halt work if the criteria are not met. If your inspection process is simply a checklist completed after the work is done, you do not have hold points. You only have a list of mistakes that were made.

It is also important to differentiate between an acceptable discontinuity and a rejectable defect. For instance, in many cases, the existence of a few pores or minor undercutting due to porosity does not automatically lead to rejection based on a code’s acceptance criteria. Replacing an acceptable weld with one that meets the same standards is a waste of time and materials. Your inspectors should be working with the same acceptance criteria as the applicable code, not their own ideas about what “looks right.”

Preheat and Interpass Temperature Aren’t Optional Extras

Controlling heat input, preheat, and interpass temperature is essential because they determine the cooling rate through the weld as well as the heat-affected zone. This, in turn, influences the grain structure and final mechanical properties. When preheat or interpass temperature is not within the range specified in the WPS, for example, hydrogen-induced cracking can occur in the weld or heat-affected zone, even hours or days after the welding is completed.

To keep these parameters in check, requirements for each welding procedure supporting a WPS must be monitored and recorded. An experienced welder can estimate the preheat temperature within a reasonable margin, for example, but to control interpass temperature it must be measured. A controlled interpass temperature is especially critical for a multiple-pass weld on one side where you cannot sense the heat with your hand. A log providing the actual preheat/interpass temperature observation and verification should be maintained.

Consumable Handling is Quality Control Before the Arc Starts

Low-hydrogen electrodes and flux-cored wires are hygroscopic, they readily absorb moisture from the atmosphere. Moisture is one of the simplest and most common sources of porosity and hydrogen cracking. This isn’t new; it’s been a known issue for more than 40 years. But controlling those variables means shop owners have to know and invest in a lot more than just power sources and welding procedures. Electrode ovens, controlled drying cycles, and documented issue-and-return procedures for consumables aren’t fussy extras for shops with money to burn. They’re a formal control point that prevents defects before welding even begins.

Traceability Turns a Weld Into a Defensible Record

On high-spec work, you should be able to reconstruct the complete history of any critical weld months after it was made. A weld map should link each joint to its WPS, the PQR that qualified it, the consumable batch used, the welder ID, and the material test certificate confirming the base metal’s chemistry and mechanical properties.

This level of traceability is what separates a certified fabricator from a general shop that does good work but can’t prove it. When a client or auditor asks “show me this weld was made correctly,” a weld map and full paper trail answers the question in minutes. Without it, you’re relying on memory and hoping nobody asks a hard question about a weld made eight months ago.

Rework is a Number, and it Should Scare You a Little

Many shops give rework the impression that it’s a problem they have without ever putting numbers on it. The American Society for Quality figures the cost of poor quality can be anywhere from 5% to 30% of gross sales for manufacturing organizations. You can apply that to a welding business, where rework and repair scrap, along with the labor to make all that right, stop being an operational annoyance and become a line item worth looking at every possible way to minimize.

Hours spent laying in repair welds, plus welding rod/tungsten wasted on repairs, multiplied by the cost of that labor and those consumables gives you a minimum rework and repair cost, and the actual cost is higher, because there’s also the labor of having to re-test NDT when a repair interferes with an inspected seam, or the added cost of missing a deadline because a joint has to be rejected and fixed.

First thing to say: “All of it.” The porosity caused by a leaking roof, the lack of fusion because 60% of your welding power sources are under 300 amps, the undersized consumables because suppliers won’t sell you the 2mm tungsten you need, all of it gets counted against the cost of poor quality. Then, once you have that number, you can make a pretty rock-solid business case for fixing lots of things. Root cause analysis on recurring defects isn’t a voodoo art where the axes must be finely tuned and the blood of a chicken sprinkled just so on the WPS. If retesting porosity at an NDT facility is soaking up your budget, and rework porosity is 35% of all repairs, and wet rod is causing porosity, it becomes a pretty clear spreadsheet what you need to do next.

Published: 2026-08-11 17:15:15