Biz Wave: The Development of Hyperbaric Pipeline Welding

Pipe Cutting Hydraulic
Ocean Systems’ divers monitoring the hydraulic pipe cutter (Christopher Swann).

This chapter is an excerpt taken from Christopher Swann’s book, “The History of Oilfield Diving“. We also have a special giveaway – a signed copy of his book! Check the details at the end of this chapter to enter.

It was said that when a pipeline went off the stern of a pipe laying barge nobody wanted to see it again for at least 25 years: a wish that storms, corrosion, pressure surges and ships’ anchors all too often confounded. Making a repair was a cumbersome business. It took two or more crane barges to lift a pipeline to the surface, and when the repaired line was set back on the bottom it often had a bend in it, which induced stresses. Welding flanges onto the cut ends and having divers insert a spool piece avoided that problem; but  flanges could leak. Clamps, installed over the damaged area by divers, were a temporary solution.

Welding a pipeline under water was out of the question. The quenching action of the water produced brittle welds that could not meet the American Petroleum Institute (API) Standard 1104, Standards for Welding Pipelines and Related Facilities.

But what if the welds were done under water in a dry chamber? What indeed?

In early 1966, Ocean Systems, collaborating with the Linde Division of Union Carbide, decided to find out. Preliminary investigation and testing revealed that the most suitable technique was tungsten inert gas welding (TIG), a procedure where the welder adds filler rod manually while maintaining an arc with a non-consumable tungsten electrode. Using TIG, the investigators made a series of pipe welds at atmospheric pressure, and at 110 feet in a chamber. Both sets of welds were x-rayed and cut into ‘coupons’, which were then machined and tested according to the requirements set out in API Standard 1104. In every case, the welds done under pressure were of comparable quality to those done at the surface, and all exceeded the API standard.

Having established the feasibility of welding at increased pressures, the Ocean Systems engineers designed a welding chamber, of which two examples were built. The Submerged Pipeline Repair System, as it was called, consisted of a central box section, which was open at the bottom and sealed around the pipe, contained within a strongback which aligned the pipe ends with large hydraulic clamping shoes. It had an overall length of 25 feet, a beam of 10 feet and a height of 8 feet. The dry weight was 29,600 pounds.

Since pipelines are usually buried, to make a repair divers had first to jet and air lift away the mud and excavate a hole for getting in and out of the chamber. Once the system was clamped to the pipeline, guided into place along down-lines, the surface operators blew the buoyancy tanks to reduce the weight. Divers then entered the flooded chamber, chipped off the cement weight coating and the composition coat and cut out the damaged section. At this point the doors were bolted around the pipe at either end and sealed, and the chamber was blown dry. To stop the chamber from exerting upward pressure on the line, the ballast tanks were then flooded.

Dry Welding Chamber
Ocean Systems welding chamber

The first pair of diver-welders, standing on gratings which were lowered down on either side of the chamber, set up the hydraulic cutter, a machine which inched its way around the pipe on a chain-strap arrangement, cutting and bevelling the ends to the required 37 ½ degree angle. Pairs of divers followed to monitor the machine and direct cooling water onto the blade. Once one stub had been cut and bevelled, the cutter was transferred to the other stub. Workers on the barge then made up a replacement section of pipe, known as a ‘pup joint’.

During welding the chamber was purged with nitrogen to eliminate the possibility of fire and prevent contamination of the weld. The divers breathed through a mask.

In September 1967, Ocean Systems undertook the first hyperbaric welding contract, a ‘hot tap’ in the Gulf of Mexico: a process in which a new line is connected to an existing line without shutting it down. In this instance, a 6-inch line was connected to a 10-inch line. (The first company to do a mechanical as opposed to a welded hot tap was S & H Sub-Water Salvage. The entire process was done in the wet using a Hydrotech hot tap saddle and a T.D. Willamson hot tap tool.)

To make the connection, Ocean Systems used a simple igloo, cut out like the larger welding chamber at either end to sit astride the pipe. The depth was 110 feet. Divers working from the surface prepared the line and set the chamber; the welding was done in saturation. As soon as the welders completed their work, the chamber was removed. Divers then carried out the hot tap in the wet with a tapping tool. The customer, Humble Pipeline Company, reduced the pressure in the pipeline during welding but resumed full operation within three days.

Hyperbaric Welding Chamber_Diver
Hot tap chamber drawing.

Between October 1967 and March 1970, Ocean Systems carried out seven further hyperbaric welding operations: a pipe fitting repair, a riser repair and three pipeline repairs in the Gulf of Mexico, a pipeline repair in the Raritan River in New Jersey, and the Halibut pipeline tie-in in the Bass Strait.

The Bass Strait contract, for Esso Australia at a depth of 220 feet, was a major undertaking, with 45°F (7°C) water and a current that ran to 1 ½ knots. Two lay barges, one working from shore to the platform, the other from the platform to shore, handled the 24-inch pipeline.

The crew consisted of pipeline welders from the Gulf of Mexico and support divers from Santa Barbara under Whitey Stefens. Dan Wilson, convinced that it made more sense to train a welder to dive than a diver to weld, had some time before persuaded the Morgan City office to hire several skilled pipeline welders and turn them into divers. In fact, the only diving the welders had to do was to go from the surface, or from the bell, to the welding chamber. The support divers removed the weight coating, cut the pipe and set the chamber, and put on the epoxy coating after the welds were x-rayed. The five welders on the Esso Australia contract spent ten days in saturation, inclusive of decompression, completing the hook-up in less than the scheduled time.

Despite these successes, lack of a heavy construction connection meant that Ocean Systems, although it had pioneered hyperbaric welding, all but ceded the field once Taylor Diving entered the market.

Offshore Platform Riser
A typical platform riser (Photo from Joe Schouest, OOGHP, 2002, a study funded by the U.S. Department of the Interior, Minerals Management Service).

Hyperbaric welding was right up Taylor’s alley. In 1961, Mark Banjavich and Jean Valz had worked on the laying of a pipeline from a tanker-loading terminal to a refinery at the Ilhad’Agua, just off Rio de Janeiro. After the lay barge left, a ship dropped an anchor on the line and put an eight-inch crack in the pipeline. With the barge gone, there was no way to pull the pipeline back to the surface; so Banjavich told the superintendent, who had stayed behind with some of the crew, that he would weld it under water. No one had ever welded a pipeline under water before.

Undaunted, Banjavich set about the repair. It took him close to a month. Day after day he descended the 60 feet to the soft mud bottom and dropped into the eight-foot deep trench to the pipeline. Visibility was virtually nil. Little by little he laid a pattern of overlapping beads over the damage, ground them smooth and ran another set of beads at right angles to the first, until the weld was three-quarters of an inch thick—the same as the wall thickness of the pipe. Finally, he welded one-inch by two-inch braces across the repair.

When the company pressure tested the pipeline, the weld held.

The repair in Brazil was the product of necessity, a performance that was unlikely to be repeated, but it set Banjavich thinking about the potential for welding pipelines underwater. Inevitably, as oil and gas companies laid ever more pipelines the chances of those lines being on the receiving end of an anchor were bound to increase. Here, surely, he told George Morrissey, was a promising business opportunity.

Underwater Welding Habitat_Taylor
The first Taylor Diving welding habitat (Collection of Drew Michel).

The company had to develop the ability to repair pipelines without pulling them to the surface. Beyond that, Banjavich saw an even more lucrative market. At greater depths, oil companies were bound to run into difficulties connecting newly laid pipelines to production platforms. The Gulf Coast method of lifting the line to the surface and cutting it to fit, then welding on an L-shaped riser and lowering it down again while trying to guide the riser into a set of clamps was not going to work. The connection would have to be made on the seabed.

The Taylor hyperbaric welding programme got under way in 1967 with the completion of the new chamber complex at Belle Chasse. The principal people involved in the development work were Morrissey and Taylor’s chief engineer Anthony Gaudiano. The welders, young, skilled and willing to try something new, came from Brown & Root. All except two, the Stockstill brothers, returned to their original jobs once they had gone offshore and found out what it was like to live in a chamber and work on the bottom in a habitat. (Lyle Stockstill later became the president and CEO of Torch Offshore in Gretna, Louisiana. The company filed for bankruptcy in 2005.)

Taylor then recruited a second group of Brown & Root welders, while simultaneously training some of its divers in hyperbaric welding.

From that experience, Taylor concluded—unlike Dan Wilson—that it was easier to train a diver to weld than train a welder to dive.

A diver could be trained to weld in about four weeks and had no reservations about climbing out of a bell in black water. Ultimately 90% of the company’s welders were divers.

Morrissey and Gaudiano began by experimenting with conventional arc welding. The results were not encouraging. Electrodes which performed well on the surface performed poorly under pressure. At 33 feet, the welders noticed a change; at 200 feet, it was rather like trying to stick hot chewing gum onto ice: the bead simply fell off the work. Cellulosic electrodes, which are very porous, did not work at all. Gas samples revealed they released large amounts of hydrogen. Of the 37 electrodes Taylor tried, only the Atom Arc, a low-hydrogen rod for welding on special pipe, proved satisfactory. As with all low-hydrogen rods, the welder started at the bottom of the pipe and worked up, rather than the other way round as with conventional rods: a technique that took some getting used to. Even with the Atom Arc, however, the welders found they could not bridge the gap in the typical fitup between two joints of pipe to put in the root pass, the critical first weld.

Welding Pipe Alignment Frame
A pipe alignment frame at Taylor’s yard in Belle Chasse, Louisiana (Don Barthelmess).

These considerations led Taylor to TIG (tungsten inert gas welding), the procedure that Ocean Systems was already using. TIG had the disadvantage that it was slow but it was the easiest method for obtaining a quality weld. Tests showed that it provided better control than with the standard arc process and it allowed for considerable errors in spacing. A skilled welder could cross a quarter-inch gap without difficulty; in fact, subsequent field experience showed it was possible to bridge an even wider gap. Like Ocean Systems, Taylor purged the chamber with nitrogen during the welding phase to eliminate the fire danger and prevent contamination of the weld. It also used argon, a very heavy gas, expelled from the torch, to shield the weld.

In the autumn of 1968, Taylor got an opportunity to put its welding technology to work. Brown & Root, supported by divers from Taylor, had pulled a bundle of small diameter pipelines across the St Lawrence River at Montreal. The lines were laid into a 30-by-30-foot trench blasted into the rock of the riverbed, then covered with ballast rock, after which inspectors from the city, the province and the Commonwealth each conducted an independent inspection. While the inspections were going on, a ship dropped one of its anchors at a point where the rock cover was inadequate. The anchor snagged two of the lines, one of which was to supply Montreal with heating oil, and pulled them up on the side of the trench. The anchor chain then broke.

By this time, the divers were back in New Orleans. Banjavich flew to Montreal, assessed the situation and told the pipeline owners that Taylor would cut out the damaged section of each line and weld in a large pup joint.

Taylor now had a group of welders certified for TIG to the API standard, but they did not have a welding habitat. Through the Brown & Root pipeline engineering department, Taylor obtained a copy of a patent for a method of welding a pipe in a flooded ditch. The drawings showed a Quonset Hut affair, with slots at either end, which sat over a pipe. Doors bolted onto the ends and sealed around the pipe with a gasket, in the manner of a packing gland. Unless the patent was amended, it had only two more years to run. The patent lawyer for Brown & Root advised that unless the inventor was aware of the potential for hyperbaric welding, the practical course would be to buy the patent for a nominal sum, or infringe it, since towards the end of a patent the holder rarely goes to the expense of defending it.

Pipe Alignment Frame Drawing
The Taylor Mod-1 pipe alignment frame (Collection of Drew Michel).

Acting on the advice, Taylor proceeded to build its first underwater welding habitat, UWH-1. Given its origins, UWH-1 was an undeniably more basic approach than that taken by Ocean Systems, lacking as it did a strongback with which to align the pipe ends. Construction, entrusted to Equitable Equipment, went on almost round the clock. The atmosphere was frantic. Everyone lent a hand, irrespective of their position in the company—Banjavich included. Such was the haste that Gaudiano was still squirting paint onto the habitat as it was loaded onto the truck for the journey to Canada.

When he arrived in Montreal Banjavich rented every big compressor he could find and rigged up two fourfoot-diameter air lifts to suck the rock and overburden out of the trench. The bottom of the trench was no more than 60–65 feet below the surface; but the damage was near mid-channel where the current was strongest, and where heavy ship traffic limited work to daylight hours. Since the available barges, designed for the river, were much smaller than those Taylor generally used, they lashed three together. Cranes on the outboard barges held the pipe off the bottom, while the welding habitat was suspended from the middle barge. Because of the current, each diver was lowered into the trench on a large concrete block, his arms and legs locked tightly round the wire.

After cutting the first of the two damaged pipelines, the divers set the habitat over the line and closed the doors underneath it. Hydraulic cylinders then drew up the doors, clamping the pipe between the jaws that made up the top of the slots in the habitat and the corresponding jaws in the doors. Being small, the pipelines were flexible enough for the divers to pull the ends into alignment with chain-binders shackled to pad-eyes in the chamber.

Offshore Pipeline Platform
Making a pipeline connection at a platform (Collection of Drew Michel).

Passing ships did not impede the work—but small boats created havoc. First one barge started rocking, then the second, then the third. If the welders were doing the root pass, the movement was enough to break the weld. They would then have to re-bevel the pipe ends and start again. Despite the difficulties, the welders ultimately produced excellent welds, which the Pittsburgh Testing Laboratory inspected by making gamma ray exposures with a Pipeliner camera.

The actual welding took about two weeks; the preparation, mobilisation and demobilisation took about three months. Including development costs, Taylor had spent $3.5 million. Their money was certainly not wasted.

By the time Banjavich left the company in 1972, Taylor dominated the hyperbaric welding business and was routinely completing welds in 72 hours, in much deeper water.

During the St Lawrence River operation, Banjavich decided that TIG was too slow and that Taylor should switch to metal inert gas welding (MIG). The MIG process differs from TIG mainly in that a wire electrode feeds into the weld automatically from a reel, by means of a ‘gun’. Banjavich instructed Gaudiano to fly back to New Orleans and look into the matter.

On the recommendation of a Taylor employee, Gaudiano obtained two Hobart MIG machines. As it turned out, the fact that the machines did not have an independent current control made them unsuitable for hyperbaric welding, although no one discovered this until later. At any rate, by the time the trucks returned from Canada with the equipment Gaudiano and his team had done a considerable amount of experimenting in the small depth simulator.

The tests showed that MIG was unquestionably faster than TIG; but the welders could not get adequate penetration at depth. To increase the heat they tried various cover gases other than argon, including blends of helium. They also ran into difficulties with the wire hanging up in the continuous feed mechanism.

Eventually, after rebuilding the MIG machines with different parts, the Taylor technicians got the results they wanted. Nonetheless, laboratory tests continued to show that the TIG hand-weld wire method, while slow, produced the highest quality welds. For some years therefore, Taylor did the root pass and the following two or three passes with TIG and completed the weld with either MIG or the Atom Arc low-hydrogen rod—although, according to Ken Wallace, the low hydrogen rods had a tendency to leave gas bubbles in the weld, which showed up on the x-ray. The weld then had to be ground out for four to six inches on either side of the defect and a new root pass put in, followed by the fillet passes. This could happen three or four times on one weld. Each grind-out increased the chances of having to do another grind-out on top. Ultimately, Taylor did the entire weld with TIG. Once a welder became proficient, there were no grind-outs.

Dry Welding Habitat
Welding habitat with jack-up legs, for use without an alignment frame (Collection of Drew Michel).

Towards the end of 1968, Taylor Diving used the welding habitat to repair a 20-inch line in the Gulf of Mexico belonging to the Tennessee Gas Pipeline Company. Again, the damage was caused by a ship’s anchor. The operation was completed in four days, of which less than four hours were taken up with welding. With the success of the Tennessee Gas and St Lawrence River repairs to point to, Banjavich organised a luncheon meeting at the Petroleum Club in Houston, to which he invited the executives and chief engineers of Brown & Root and a number of major oil companies. The main topic of conversation was setting platform risers in deep water. The oil company engineers clearly thought it was something they could handle: Banjavich had to show them they were wrong. He got up from the table and invited them onto the balcony, 23 stories up.

‘I said, “One of you guys go down there and hold a teacup, and I’ll give another of you a piece of piano wire. In this wind, you try to stick the wire in that teacup.” That was all it took. I think it was the guy from either Esso or Texaco who said, “Okay, you made your point!”

‘It was not pre-planned; I was sitting there thinking about the currents and the tides and the rough water, and the barge jumping up and down. I realised I needed a way to demonstrate the problem to them. There was not one dummy in that crowd; I had to convince them. I realised if I only half-convinced them it wouldn’t fly.

‘The point I was making was that if you want to go up to a structure in, say, 240 feet of water—of course we worked beyond that—and you hold a pipeline up near it, you can visualise the long catenary. You’ve got to suspend that pipe out a couple of thousand feet so you don’t bend it, then set it down so it sits in the 20 saddles up along the diagonal jacket leg. It’s an impossibility! So I told them, “What you do is set the riser in place when you set the structure. All we do is lay up to it, then go down and cut it and make an effective underwater weld”—which we did eventually.’

Banjavich already had an oral agreement with Brown & Root to give Taylor the money to build a pipe alignment frame; the pipeline division even set its engineers to designing it. The frame had to be able to move the pipe horizontally and vertically, and it had to be very strong. Banjavich and the Taylor engineers thought the ends of the pipe should sit in jaws which would move on a metal slab that would be left on the bottom.

Welding Dive Saturation
A Taylor welding spread and saturation system on a Brown & Root derrick barge (Collection of Drew

The purists at Brown & Root did not want to leave the slab on the bottom; they designed a gigantic frame with 60-inch-deep beams that sat athwart the pipe. This was not at all what Banjavich had in mind, so he and his engineers designed an alignment frame of their own. However, instead of using tubular sections as Gaudiano advised, Banjavich opted for structural members, which made the frame too heavy. Consequently, on its first outing it sank into the mud: a shortcoming that was overcome by adding buoyancy tanks.

In 1969, Taylor began welding risers intensively. In 1969–70 the company did the majority of the hyperbaric welds they were to make in the Gulf of Mexico, using a saturation system on a small barge. Most were done without the alignment frame because, contrary to what Banjavich had asked for, Brown & Root put a 15-foot stub on the ‘L’ of the riser, not a 30-foot stub. The shorter stub left insufficient room. To make the alignment without the frame, Taylor fitted the welding habitat with jack-up legs and added hydraulic cylinders at either end to move the pipe horizontally and vertically. With the pipe in the jaws and the habitat flooded to the top of the slots, two divers, one inside and one outside, could align the pipe ends with comparative ease.

Soon Taylor was completing riser connections in three days of work on the bottom: a very short time considering that it took Brown & Root three to four weeks with the conventional method.

‘A considerable factor in our being able to do those jobs so fast was the co-operation of the people who put in the riser and laid the pipeline,’ recalled Gaudiano. ‘But in the beginning the co-operation wasn’t too good. Those guys were a bunch of cowboys. We’d find that the riser would be in the clamps you could see, but on the bottom, it would only have been installed in half of them. As a result, we’d have to spend some time getting the riser in the clamps. Then they were supposed to lay the pipe alongside the horizontal leg of the riser and sometimes it would be a quarter of a mile away. Finally, after numerous meetings with Brown & Root, they got their superintendents together and told them, “Look, this is how you’ve got to do it, or else.” So then everybody started working together.’

Hyperbaric Welding Habitat_McDermott
McDermott welding habitat and alignment frame (J. Ray McDermott & Co. Inc.).

The other giant of the offshore construction business, J. Ray McDermott, got into hyperbaric welding when they buckled a large-diameter gas pipeline in 200 feet of water. Because the damage was done during laying, it was up to McDermott to splice in a new section at their own expense. After spending over $1 million trying to line up and connect the ends using flanges and various other methods, McDermott built a massive 165-foot long alignment frame, with a 20 foot by 10 foot welding chamber that lowered into place in the middle of the frame. The construction costs ran to approximately $2 million. This was the forerunner of even larger alignment frames built by both Taylor and McDermott, designed to grip the pipe over a sufficiently long span to eliminate any risk of bending or kinking.

McDermott followed essentially the same procedure as Ocean Systems and Taylor. Dick Evans Inc. certified four welders—two divers trained to weld, two welders trained to dive—to the required API standard and completed the work in three or four days (having planned on two weeks). The system was subsequently used for several other repairs. Dick Evans recalled the operating cost being $20,000–30,000 a day, although even at that level it apparently paid for itself.

Despite their size and position in the industry, McDermott never mounted a serious challenge to Taylor Diving in the hyperbaric welding arena, either in the Gulf of Mexico or overseas. Until Comex exploited an opening through the French oil company Total, Taylor had the market virtually to themselves.


Christopher is giving away 1 free, signed copy of “The History of Oilfield Diving“.

Rules: To enter into the giveaway, include a comment or question in the comment section below related to Christopher Swann and his book. One winner will be chosen in a drawing and contacted.

Deadline: August 3, 11:59 pm (CST).

5 thoughts on “Biz Wave: The Development of Hyperbaric Pipeline Welding”

  1. from the Tv I know red adair and other people that developed firefighting in the oil industry, but a Tv special should be made for these men in underwater welding that don´t work as exposed to the media. They were pioneers

    • They absolutely were, Carlos! They’ve paved the way for many commercial divers and industries after them to improve upon technology and processes underwater.

    • Andrea:

      Thanks for your comment! You’ve been selected as the winner of one free copy of Christopher’s book, “The History of Oilfield Diving”!

      Can you email me your mailing address so he can send you it? You can message me at

      Thank you for participating!

Comments are closed.