How one company overcame major challenges to achieve quality results
July 15, 2010
Autogenous orbital welding on large projects has become nearly routine with large numbers of successful welds completed with very low reject rates. Thus it is very easy to become complacent. However, installers need to be alert to potential problems and take immediate action when a problem occurs.
Since the early 1980s autogenous orbital gas tungsten arc welding (GTAW) has been the preferred technology for joining of high-purity process piping for bioprocess, semiconductor, and other high-purity applications. Over time millions of quality orbital welds have been made, but new challenges arise even with the widespread systemization of welding approaches.
Welding of the predominant 0.25-, 0.375-, and 0.50-in.-OD tube sizes in the semiconductor industry and the 1- to 4-in.-OD tube sizes used for bioprocessing equipment has become routine. However, welders in both industries occasionally need to weld 6-in.-dia. tubing and fittings with a wall thickness of 0.109 in., which is at the upper limit of diameter (but not wall thickness) for autogenous orbital welding with enclosed weld heads.
Because this size is near the edge of the process, orbital welding becomes trickier. The tube's ovality can make it difficult to maintain a constant arc gap with a fixed tungsten length, which is needed for weld repeatability. Material differences between the tube and the fitting can present problems during welding. The location of these welds also may present access problems.
Kenyon Engineering (see Figure 1), a mechanical contractor in Singapore, recently encountered a problem with autogenous orbital GTAW of 6-in.-dia. tubing-to-T fittings. The fittings were for REC Solar, a company that specializes in designing and installing grid-tier solar photovoltaic systems, and the company's $6.1 billion solar panel production project, in which Kenyon played a very small part. Kenyon reported that all of its weld problems occurred at T joints done at high levels on a pipe rack. An investigation uncovered several reasons why this joint in this location presented a problem.
Arc Gap Variations. Enclosed weld heads for autogenous welding rely on a tungsten electrode set in the rotor of the weld head that rotates around the weld joint to complete the weld. (The head does not rotate, only the rotor.) The arc gap—the distance between the electrode tip and the tube surface— determines the arc voltage, while the power supply maintains a constant current level set by the weld program.
With a perfectly round tube, the arc gap is constant around the entire tube circumference. However, it is difficult to maintain the circular shape of 6-in.-dia., thin-wall tubing through shipping, handling, and storage, even if it met industry ovality standards at the time of manufacture.
If the arc gap varies as the electrode rotates around the tube, arc voltage and heat input become inconsistent. This can result in variable penetration of the weld joint and lack-of-penetration defects. If the arc gap is too tight, the tip of the electrode may contact the weld pool, causing the arc to stub out.
Arc Machines Inc., the manufacturer of the orbital welding equipment Kenyon used for this project, recommended an arc gap of 0.070 in. for welding the 6-in. tube. This was determined by the length of the tungsten electrode installed in the rotor.
Kenyon developed a special clamp that attaches to the tube ends near the joint (see Figure 2). Four separate screws were used to adjust the circularity of the tube circumference before tack welding. This device helped to ensure a more uniform arc gap, but some problems still remained.
For example, even if the roundness is corrected, the OD/ID of large-diameter tubes can vary substantially within the industrial tolerance limits. Sometimes such variance is so big that it is difficult to weld them together. As a good practice, it is important to check and record the OD/ID distribution of tubes at the time the material is received, then sort and match them according to the nearest dimension, and plan the material flow accordingly. Otherwise, toward the end of a project, the difference of the OD/ID of the remaining tubes is simply too big to be welded together. As delivery time for large-diameter tubes can be very long, the project schedule can be badly affected if sorting precaution is not taken early.
Alignment and Fit-up. Autogenous orbital welds are done with a square butt joint. The fit-up must be tight, with no visible gap between the tubes. If a gap exists, the arc may deviate to one side or the other rather than consume the joint. Installers sometimes can eliminate the gap by turning the tubes to achieve the best fit.
The American Society of Mechanical Engineers' (ASME's) Bioprocessing Equipment (BPE) standard specifies material chemistry and dimensional tolerances for weld ends of fittings, valves, and other process components. Even when fittings are within these tolerances, it is possible to get unmatched wall thicknesses with the larger 4- and 6-in. tube sizes. Tolerances can stack up and appear to fail the BPE specifications for misalignment: 15 percent of the wall thickness for tubing less than 4 in. OD, 0.015 in. for 4-in.-OD tube, and 0.030 in. for 6-in.-OD tube.
If the tube is not perfectly square on the ends, installers sometimes can eliminate a small gap by turning the tube to achieve the best fit. When joining tubes to fittings with the potential for misalignment, contractors typically position the components concentrically for a vertical orientation or, when horizontal or some degree off vertical, match the ID surfaces to allow for drainability.
Kenyon installers routinely rotated the tube-to-tube welds in the pipe rack at ground level, where they could be adjusted for best fit-up and checked. However, rotating the tubing when it was high in the pipe rack proved difficult, and checking the fit-up was not as easy as doing it at a lower level. As some of the 6-in. fittings were delivered late, rotation of the tubing was not possible at all, and this was the source of the greatest problems.
As a result of that experience, Kenyon stopped tack welding the tubing in place before the arrival of fittings because it was very important to be able to rotate the tubing to get a good fit-up.
Tack Welding After the Fittings. Orbital weld heads are not designed to support the weight of long lengths of tubing to hold them in position. If larger-diameter tubes are not tack-welded in position before welding, even with good fit-up the components tend to expand during welding, creating a gap that cannot be consumed by the arc.
Kenyon found that a few of the problems on the 6-in.-OD tube were caused by overtacking—the current was too high—and insufficient purging. Tack welds must be strong enough to hold the tube ends in place, but not so large that they will not be consumed by the orbital arc as required by the ASME BPE standard.
Purging the inside and outside of the weld joint with inert gas during tack welding is critical. Insufficient purging of tacks can cause the orbital weld bead to deviate around the tack, resulting in lack of penetration. Excessive oxidation on the ID or OD of the tack weld can result in failure to consume the tack (see Figure 3).
Kenyon discovered that the arc typically favored the higher side when it passed the tack weld. To prevent this, operators typically filed a very shallow groove in the direction of the weld seam across the tack weld. The nearly invisible groove centered the arc in a straight line, resulting in a better weld appearance.
In addition, Kenyon used the step travel mode for welds on wall thicknesses 0.12 in. or greater. Welding in step mode is comparable to a manual welder walking the cup. Rotation of the rotor with the tungsten electrode is stopped during the high-current pulse to provide maximum penetration, then moved during the low-current pulse. This technique provided a wider weld bead with more penetration achieved at the same current level. The arc time was about twice that for continuous welding.
While a step weld might not be as attractive as a continuous weld, it more readily consumes tack welds, handles mismatch, and helps to minimize weld pool shift when welding high- to low-sulfur heats of material. Kenyon preferred to limit the use of step mode to welds on wall thicknesses of 0.12 in. or greater.
Overheated Weld Heads. Some type of water cooling is fairly critical for orbital welding of the 4- to 6-in. tubing. For smaller-diameter, thin-wall tubing with air-cooled heads, if the weld head is allowed to cool sufficiently between welds, the weld usually can be completed before excessive heat can build up in the weld head.
Admittedly, allowing time for cooling has an adverse effect on productivity. For larger diameters, the head is likely to become overheated before completing the weld. A water cooling unit is an essential accessory for welding these larger sizes.
Training of Welding Personnel. Welding personnel must be alert to potential problems such as poor fit-up and improper purging techniques for tack and orbital welding. This is particularly critical for welds in difficult locations and with tubing sizes that are not commonly used.
Kenyon had all the tools necessary to achieve the proper fit-up, but this was not always realized when the conditions varied from the usual. Kenyon instructed its welding operators in techniques for achieving proper fit-up and for making tack welds that were strong enough to hold the tubing in place, but well-purged so the tacks would be consumed by the orbital welding arc. Their technique of making a slight groove on the tack weld facilitated the arc moving in a straight line across the tack.
Kenyon also experienced problems because the standard operating procedures were different from typical biopharmaceutical installations. All of Kenyon's welding for pharmaceutical and biotech is 100 percent inspected on the ID with a borescope. This practice gives production the opportunity to correct welds that do not meet quality standards. While the REC welds were 100 percent visually inspected on the OD, they were not initially inspected on the ID with a borescope.
As a result of the OD visual inspection, Kenyon decided that some of the welds had very bad fit-up. Those welds were cut out, and new welds made. At that point, based on visual inspection, Kenyon elected to reweld all the other joints in the line, even though they appeared to be good from the outside. The ASME BPE standard permits one attempt at rewelding for lack of penetration, so if full penetration was achieved by a single attempt at rewelding, this would be acceptable. All subsequent repair welds were fully inspected on the ID with a borescope and found to be acceptable for the REC application, which was somewhat less critical than for biotech or semiconductor process gas lines.
It should be noted that autogenous orbital butt welds for 6 in. OD up to 12-in. Schedule 10 have been done successfully for high-purity semiconductor applications with weld heads designed for orbital pipe welding. These heads typically have an arc length control function that electronically maintains a constant arc gap. This equipment is not typically used for 6-in.-dia. pharmaceutical tubing, but it has been done.
Even with arc length control technology, it is important to have adequate end preparation, excellent fit-up, good tack welds, and proven inert gas purging procedures.
Welders not involved in the semiconductor or biopharmaceutical industries should be aware that high-purity welding may be expanding into their work worlds. The ASME B31.3 Process Piping Code recently approved a new Chapter X, High Purity Piping, for the 2010 edition, to be published later this year. Chapter X pertains to piping designated by the owner as being in high-purity fluid service, defined as a fluid service that requires alternative methods of fabrication, inspection, examination, and testing not covered elsewhere in the Code with the intent to produce a controlled level of cleanness. The term applies to piping systems defined for other purposes as high-purity, ultra-high-purity, hygienic, or aseptic.
This chapter is intended to close the link between piping safety as specified in the B31.3 code and the needs of the high-purity industries for very high levels of cleanness in piping systems so they don't add contaminants to the products passing through the piping. The intent is to apply to a broader industrial base than just semiconductor and biotech and to include industries such as chemical, biofuel, dairy, and others with a need for high-purity piping systems.
Autogenous orbital welding on large projects has become nearly routine, with many successful welds completed with very low reject rates. Thus, it is very easy to become complacent.
However, installers, especially those working with larger-diameter tubing in high-purity applications, need to be alert to potential problems, pay close attention to detail, and take immediate action when problems occur.
Orbital welding by definition is "automatic or machine welding of tubes or pipe in place with the electrode rotating (or orbiting) around the work."
Orbital welding is a mechanized version of gas tungsten arc welding and, as such, can be done with or without the addition of filler wire. For high-purity applications, orbital welding is a fusion, or autogenous, process in which the squared, machined tube ends are fused together by the heat of the welding arc.
Orbital welding with filler wire is considered to be machine welding and requires some operator intervention during the weld. Autogenous orbital welding is considered to be fully automatic—once the operator initiates the start sequence, the weld is completed without further operator intervention.
It should be emphasized that while it is true that orbital equipment does not require the same skills as manual welding, much of the experience of a skilled manual welder is applicable to orbital welding. And, of course, many, if not most, operators of orbital equipment have a background in manual welding. The industries that specify orbital welding have these characteristics in common: a requirement for ID weld bead smoothness on the product contact surface, consistent full-penetration welds, and thousands of joints in tubes of the same OD and wall thickness.