The Immersed-Tunnel Method
FORGET ABOUT SANDHOGS, DYNAMITE, and compressed air. With this technology, underwater tunneling is less like digging a mine and more like docking the space shuttle.
BUILDING ANY TUNNEL CAN PRESENT A HOST OF difficulties, from getting soil out and construction materials in to performing precision alignment deep inside the earth. When the tunneling takes place underwater, a whole new set of obstacles arises.
Starting early in this century, however, advances in structural steel and reinforced concrete technology have allowed the development, mostly in the United States, of the immersed-tunnel method, in which prebuilt sections the length of a city block are floated into place and sunk to their proper position. The process is similar (on a grand scale) to the method used on land in which builders simply dig a trench, lay pipe, and cover it up. The need to keep everything in place beneath dozens of feet of constantly shifting water and mud complicates the underwater version greatly, but with a carefully crafted design and an experienced construction team, it is often safer, cheaper, and more reliable than any other way. In my four decades as a civil engineer, I have worked on a number of tunnels built by the immersed method, and my growing familiarity with the technique has only deepened my appreciation of the planning, coordination, and teamwork that go into it.
THE WINDOWED ROOM ATOP THE two-story control tower of the lay barge is dark, except for a few lights shining on the instrument panels and the glow from a computer screen. City lights waver on the harbor outside as an occasional vessel slides by. While the immediate scene below is floodlighted, it is otherwise a dark, overcast, rather chilly night. Half a dozen men sit quietly in the half-darkness listening to radio call-outs from the survey crews as a 30,000-ton tunnel section, hanging deep underwater, is very slowly and carefully maneuvered into position against the end of the previous tube. From time to time the operator pushes and pulls a handle to start and stop the travel of an anchor line. The roar of large diesel hoists alternates with the hiss of compressed-air brakes.
The superintendent calls, “All stop!” On the speaker box, the surveyor reports by radio the exact position of the forward and aft targets: “North is one-tenth east; south is four-hundredths of a foot west; coming on line.” The tube is hanging on four sets of lowering “falls,” each made up of fifteen to twenty strands of two-inch cable. The falls are carried on a pair of twelve-foot-high girders located fore and aft of the lay barge, and each one supports the equivalent of a Boeing 747. Four large dials above the windows indicate the loads on the falls. The two inboard dials read zero, showing that the end of the tube closer to shore is now sitting on the guide plates of the tunnel. The two on the outboard end read 600,000 pounds each, showing that the other end of the tube is hanging just off the bottom. The superintendent calls for the diver, and things quiet down for a while.
This tube placement started in the early-morning hours, when the hundred-yard-long, four-lane highway tunnel section—ballasted to 4 percent of its total displacement, or about 1,200 tons—was towed away from its dock. When the tides reached their optimum stage, tugboats moved the lay barge and tube along the main shipping channel, which was wide and deep enough for them to maneuver easily. After aligning with the tunnel trench, which is at right angles to the shipping channel, the lay barge was then slowly pulled and guided along by means of anchor lines.
This laborious process took all day. Since the trench was narrow, and the water outside it was not deep enough for the tube’s forty-foot draft, precision was absolutely essential. One false move of the lay barge or one error on the part of the surveyors would have resulted in grounding and damage, possibly requiring dry-docking of the $10-million tube. That would have delayed the project for weeks and cost hundreds of thousands of dollars.
Now the diver is down. His voice, hard to understand, alternates with the loud hiss of his breathing. Through a speaker in the cab, he reports that the tube is sitting hard on the shims and is lined up properly with the tunnel. “Extend the jacks!” is radioed to a console operator on the lay barge. As the diver watches, four 75-ton hydraulic jacks mounted on the mating end of the tube are extended, one at a time. These jacks will be used to pull the new tube tightly against the tunnel portion already in place. Each jack has a standard railroad-car coupler mounted on its piston, to engage with a matching coupler on the end of the tunnel. (Railroad couplers are used since they will engage despite the inevitable slight misalignment.) The diver reports to ensure that each pair engages securely.
The tube section was sealed before launching with heavy steel bulkheads at both ends, so the inside is never filled with water, and was made to sink by loading concrete ballast on the outside. Its inboard (mating) end has rubber gaskets all around the two roadway bores. When the jacks are retracted after engaging the couplers, the ,gaskets are partially compressed. This traps water in the joint area at full ambient pressure.
When the diver reports that all is well, the superintendent tells a crew inside the tunnel to open valves in the two bulkheads and pump the water out of the joint space back into the harbor. Pressure in the joint drops to atmospheric, while full water pressure prevails at the other end of the tube. The resulting unbalanced force moves the gigantic tube forward a couple of inches to close the joint tightly against the rubber gaskets.
When equilibrium is reached, the gaskets will be under compression with a force equal to several thousand tons. This will provide a virtually watertight joint between the tube just placed and the tunnel already built. Work crews will eventually construct the roadway and walls of the tunnel through the joint area after the steel bulkheads have been removed. Before then, however, one more tube must be added, so that three watertight bulkheads will always be in place to protect the tunnel against flooding. Also, the joint area will be cleaned of rust, mud, and decaying marine growth and then have a closure plate installed across it to provide a permanent, continuous water barrier between the tubes.
But none of this can happen unless the seal is made absolutely tight. Back in the cab on the lay barge, the command to open the valves has just been given. The survey team reads off the distance, measured electronically, from precisely located markers on land to a target prism mounted on the survey tower. The tube moves slowly as the pressure drops in the joint.
The survey tower is a temporary steel framework, about fifteen feet on a side, mounted on the outboard end of the tube being placed, and tall enough to extend some twenty feet above the water after the tube is on the bottom. A ten-inch pipe down the center of the survey tower runs vertically to the top of the tube. At the bottom of this pipe, which is kept dry inside, is an illuminated survey target. An instrument on top of the pipe sights a perfect vertical line to the target, and it transmits the exact position of the tube from the bottom of the harbor to the top of the survey tower, where it can be located from onshore reference marks to an accuracy of an eighth of an inch.
The final step is a precise check of the tube’s alignment, elevation, and rotation about the longitudinal axis. The position tolerance is generally specified as a maximum of one and a half inches in any direction. If the alignment is not satisfactory, minor adjustments can be made by flooding the joint and resetting wedges to correct the horizontal line. If there’s a problem with rotation or elevation, however, the whole operation may have to be aborted, with the tube being disconnected and moved out of the trench so that the gravel foundation course can be fixed. This is very costly, so it pays the contractor to get it right the first time.
As dawn begins to break, the engineer accepts the position and condition of the tube and signs off on the survey data sheet. The contractor immediately starts disconnecting the lowering falls and pouring concrete through an underwater pipe to fill the remaining ballast pockets and bring the tube to the specified negative buoyancy of about 2,000 tons. Weary crews from the previous day and night board the tug and head for home. The tube has been set to within a quarter-inch of line and a half-inch of grade, with no measurable rotation. Everybody is on the high that comes from having contributed to doing a demanding job well. The tunnel is a hundred yards longer than it was yesterday.
I ENCOUNTERED THE IMMERSED -tunnel method for the first time as a newly graduated structural designer in the mid-1950s. I remember chuckling to myself: “They plan to hook up sections of tunnel underwater? They must be crazy! This I have to see.” I was working at the time with Donald N. Tanner, a fine engineer who had already completed the design and construction of a couple of these tunnels. Don told me not to worry, that by the time our new project was over we would put twenty-three tubes together and meet in the middle of Hampton Roads. I did not realize at the time that this type of work would eventually involve the greater part of my career. (Don passed away at his desk in 1984 while working on his ninth tunnel.)
The original concept of the sunkentube tunnel was devised by two very ingenious British engineers: Charles Wyatt, who proposed the idea in 1810 in a competition to build a tunnel under the Thames River, and John Isaac Hawkins, who carried out a largescale test of the method. Two brick cylinders with domed ends, having an internal diameter of nine feet and a length of 25 feet, were constructed on a barge. The barge was then sunk out from under the cylinders, and after floating into position, the cylinders were sunk and joined together underwater. The test was successful, but far more costly than had been expected, and the client became discouraged and soon abandoned the idea. In light of what we know now, it is probably fortunate for the workers that the scheme was not carried out. The method had to await the development of reinforced concrete and structural steel early in this century to be viable.
The concept of concrete reinforced with steel bars was patented as early as 1857 and saw some use by the 187Os. Its properties were not well understood, however, and only in the early 190Os, when reliable standards became available, did it start to be widely employed. Around the same time, structural steel was also coming into its own. Before then, brick and stone masonry had been used for most structures, including the lining of tunnels.
Masonry is adequate for tunneling through a mountain or in an excavation, where the loadings are radial and generally constant. However, submerged tunnels also have significant bending loads, which change constantly during construction, ballasting, and placement. Masonry could not safely handle these stresses. It took concrete reinforced to withstand bending loads, along with structural steel to provide water tightness and share the loadings, before immersed tunneling could be made practical.
The first immersed “tunnel” successfully completed was actually a sixfoot sewer main, the Shirley Gut Siphon, in Boston, in 1893. The first full-scale tunnel using the method was constructed under the Detroit River in 1910. It was for a two-track railway and comprised eleven tubes, each about 260 feet long, reaching a total length of one-half mile.
People commonly think that building a tunnel under a river must involve digging—either by drilling holes and blasting with dynamite or by jacking a shield forward in soft ground. Both methods remain in wide use, and their characteristic hazards—unstable ground, flooding, and caisson disease (the bends)—have been greatly reduced in recent years. Still, accidents do happen and problems do occur. When a sophisticated, electronically controlled tunnel boring machine is flooded, for example—as happened recently on two separate world-class bored-tunnel projects—the delays can be major. It means dealing with a 30-foot-diameter hole with a huge broken drill bit stuck in it.
The immersed-tunnel method avoids these difficulties by building the tunnel aboveground in a quality-controlled environment. As described earlier, it’s not quite as simple as snapping Lego blocks together, but the construction risk—both human and financial—is much less than with other methods, and scheduling and cost are much more predictable. There is a further advantage: the reduced depth to which the tunnel must extend. Since the tunnel is laid in a trench dredged on the harbor’s bottom, engineers need only select the thickness they want for protective earth cover over the top of the structure—commonly five to ten feet, to protect the tunnel against damage from ships’ anchors. In some cases where shipping is not a hazard, the tunnel can actually be left without cover and can match or even extend above the bottom of the waterway.
A bored tunnel, on the other hand, must be deep enough to prevent a blowout (when air explodes to the outside) or the possibility of a failure and collapse. To preclude these dangers, it must be dug at least one diameter below the harbor bottom. This means that the bored tunnel has to be thirty or forty feet deeper than an immersed tunnel in the same place. This added depth demands at least an extra thousand feet at each end to get the roadway back to the surface. In a dense urban area this extra room is often unavailable or prohibitively expensive.
For these reasons use of the immersed-tunnel method has grown in acceptance worldwide, especially since World War II. Before the war only six rail and highway tunnels had been built this way, all of them in the United States. The total today stands at just over one hundred.
IN EUROPE THE METHOD WAS SLOW er to gain acceptance. Just before Germany overran the Netherlands, a roadway tunnel under the Nieuwe Maas River in Rotterdam was started, the first immersed highway tunnel outside the United States. The second tube was placed the day the Germans invaded, May 10, 1940. Five tubes were still under construction and a sixth was ready for the next placement.
When the harbor came under enemy fire, engineers and contractors fled the site after doing all they could to provide for the safety of the tubes. Machine-gun bullets punched holes in some of the tube bulkheads, and when the fighting stopped five days later, one of the tubes had sunk almost completely. Work was resumed during the occupation, and the tunnel finally opened in February 1942. At the time there was no celebration and little traffic. After the war it opened a second time, on May 19, 1945, once the Dutch removed explosive charges that the Germans had placed in it.
The “concrete box” type was born with this tunnel and has remained the standard method used in Europe ever since. This design uses a rectangular concrete tunnel structure with heavily reinforced walls and roof and floor slabs. The concrete box sections are usually built several at a time in a huge excavated basin adjacent to a waterway. After a batch of tubes has been completed, the basin is flooded and a passage is opened to the waterway so that the tubes can be towed to the tunnel site. The concrete tubes float with just a few inches of freeboard and are sunk by filling internal ballast tanks with water. Eventually permanent ballast is added when the interior roadway is paved.
THE AMERICAN TYPE OF TUBE, on the other hand, has evolved as a composite steel and reinforced-concrete ring—like a round concrete pipe within an exterior steel shell. It is constructed on shipways and then launched and towed to the project, either floating or on a barge. The concrete interior is then completed (“outfitted”) with the tube floating at the pier. Finally, the tube is prepared for sinking by being ballasted with external concrete fill placed in large pockets along the sides.
With the European method a number of tubes are made available for placement all at once, with a gap of several months between float-outs. A project with ten tubes might involve two float-outs of five apiece, spaced six months apart. With the American method the tube fabrication, outfitting, and sinking proceed more regularly and can be scheduled on a monthly cycle or even faster.
Each method has its advantages. For example, building concrete box tubes involves fairly straightforward construction, while steel tubes require a skilled shipyard operation. But excavating a large casting basin and keeping it pumped dry is costly and causes environmental problems because it depletes the water table. In the Netherlands, where many immersed tunnels have been built, a single basin has been reused for a number of tunnels. While concrete tubes draw some twenty-five feet when towed, steel-shell tubes require only a few feet of water to get to the tunnel site. They can be towed at very shallow draft for thousands of miles if necessary. Once near the site, the tunnel tubes can be outfitted at a pier and moved a short distance at full draft before being placed.
I have had the pleasure of working in the field on three separate projects of this type. First was the Second Hampton Roads Bridge-Tunnel in Norfolk, Virginia, in 1970-74. On that job I was fortunate to work with a worldclass mentor, Roger B. Stevenson, who had just finished building the longest immersed tunnel in the world, the BART Tunnel under San Francisco Bay. I also worked on the Fort McHenry Tunnel in 1980-84, and most recently, the Ted Williams Tunnel in Boston, 1992-94. The most exciting of these projects for me was the Fort McHenry. This tunnel, which completed Interstate 95 through Baltimore, was designed to carry eight lanes of traffic across the harbor under the Patapsco River without harming the Fort McHenry National Shrine, where the battle that inspired “The Star-Spangled Banner” took place.
The Fort McHenry Tunnel is the largest of its kind in the world. It required a trench 180 feet wide at the bottom. Two separate four-lane immersed tunnels were constructed in the trench with less than ten feet of space in between. To complicate matters, the tunnels had to be built on long horizontal curves to avoid the fort.
A tunnel was not the least expensive choice by any means. It ended up costing three-quarters of a billion dollars—below the original estimate, but far more than a bridge would have cost. However, the result could not have been better. While a bridge would have been too visually intrusive, visitors to the fort are completely unaware that eight lanes of highway traffic are passing only a few hundred yards away.
Once a tunnel was decided on, the need to dispose of some 3.5 million cubic yards of dredge spoil became a make-or-break environmental obstacle. A great deal of the material in the upper five to ten feet of the harbor bottom was contaminated from centuries of industrial discharge, so it couldn’t be dumped in a place where it might leak into the surrounding area. As it happened, just when a suitable disposal site was being sought, the nearby Port of Dundalk, Maryland, was looking for a place to expand its cargo container operations. Principals from the various agencies involved formed an environmental task force that worked closely with the Maryland State Highway Administration, Interstate Division for Baltimore City, and the designers. They found a suitable area of marsh and open water where the dredge spoil could be pumped, reclaiming land for Dundalk even as it provided a disposal area. To compensate for the marshland that was filled, a larger marshland was built adjacent to the historical park, where it could be enjoyed by visitors to the fort.
This convenient solution would work only if the contaminated spoil could be safely contained. Eventually project engineers came up with a unique and unprecedented method. Two large waterfront areas totaling 146 acres were enclosed with cellular cofferdams interlocked to form walls some sixty-two feet thick. One area was designed to contain the soft, contaminated bottom material; it was made watertight and sealed with clay. The other area was for more granular, uncontaminated materials.
A two-and-a-half-mile pipeline carried spoil from the tunnel trench to the disposal site in the form of a slurry of soil mixed with water. The soil settled in the basins, and the water was decanted to flow back into the harbor, moving slowly through a sevenhundred-foot-long settlement pool after chemical treatment to remove fine suspended clays and silts. This disposal facility worked well and is now the site of a busy annex for the Port of Dundalk’s container operation.
THE $426-MILLION BID FOR THE tunnel’s immersed portion was the largest single contract in the history of the interstate highway system when it was awarded in May 1980. The contract went to a joint venture of Peter Kiewit Sons’, Inc., Raymond International Inc., and the Tidewater Construction Company, using designs prepared by the Sverdrup Corporation and my firm, Parsons Brinckerhoff Quade and Douglas, Inc. Among them, these organizations had the bulk of the immersed-tunnel design and construction experience in the United States. It took $10 million and nearly a year to mobilize a shipyard to build the thirty-two tubes that would be needed. The Wiley Manufacturing Company shipyard, on the north bank of the Susquehanna River at Port Deposit, Maryland (some forty miles north of Baltimore), dedicated its entire plant to the project for its duration. Production gradually grew to the point where a tube was launched every two weeks.
Once afloat, the tubes were towed down the Chesapeake Bay, riding high in the water. Along the way they had to be threaded between the piers of the Amtrak bridge over the Susquehanna at Havre de Grace. Amtrak was understandably afraid of panic among the passengers who spotted a three-hundred-foot tube bearing down on the bridge they were crossing, and the railroad was warned each time a tube was coming so that it could hold back its trains. In fact, with only six feet of clearance for the curved tubes, the bridge piers did get scraped a few times.
On reaching Baltimore, the tube was docked at leased piers for outfitting. Some concrete had already been placed in the bottom of the tubes to stiffen and stabilize them; now, at the outfitting pier, construction workers poured almost 25,000 tons more to form the interior walls and roadways of the tunnel in the two side-by-side bores. As concrete went in, the tube settled lower in the water until finally it had only a few feet of freeboard. At this point the tube was “buttoned up”- made watertight for the sinking operation. No one would see its interior again for several weeks, until it was in its final location on the river bottom.
Before the Fort McHenry tubes could be placed, we had to prepare the bottom of the trench. In the United States this is usually done by the “screeding” method. The first step is to dredge the trench at least two feet deeper than the bottom elevation of the tunnel. A specialized rig then places gravel and sand in the bottom of the trench and smooths it. This “foundation course” has to provide uniform support for the tube at an exact elevation and slope. For uniformity, the material is placed on the bottom in a carefully controlled grid pattern and then “screeded” with a sixty-foot-wide beam operated from the rig.
The screed rig is an ingeniously designed piece of equipment capable of trimming the gravel layer to a precision of about half an inch in water depths of well over a hundred feet. That’s a little like smoothing a sidewalk from the top of a ten-story building in the dark. Before this floating rig goes to work, surveyors set it to an exact plane. From then on, adjustable anchor blocks at each corner hold it against variations in tide level. A motorized bridgework, riding on rails adjusted to the proper gradient, carries the beam back and forth under the rig, smoothing the gravel. The moment of truth for a screeding crew comes when a tube first sits on the foundation and its elevation is determined. If it agrees with the specifications, everybody is able to breathe again. The Europeans use a different system to establish the foundation. The outboard end of a new section is set on jacks to its exact position, using concrete pads on the bottom to support it temporarily while the other end is being attached to the existing tunnel; then a slurry of sand and gravel is pumped underneath for permanent support.
The Fort McHenry Tunnel was unusual because it was really two tunnels, assembled simultaneously side by side in a single trench. One result was that tubes had to be added alternately, first to one tunnel and then to the other. That’s because once a tube is placed, it has to be fixed in position with gran- ular “locking fill” to keep it from accidentally shifting off line. Since the tubes are ultimately ballasted to only about 7 percent of their displacement, they are quite vulnerable to being moved sideways. At Fort McHenry both adjacent tubes had to be put in place before locking.
Even when sideways deviations do occur, they can sometimes be dealt with. At one point during construction, for example, a bargeload of gravel accidentally overturned before the locking fill had been installed. All the spilled material settled on one side of a tube, moving it off line by some three inches at its outboard end. There was no practicable way to realign the fully ballasted tube, so we gradually took out the deviation in the alignment of the subsequent tubes. Three inches in 330 feet is not noticeable to a driver and had no effect on the watertight integrity of the tunnel.
As the tubes extended into the Patapsco River, work inside the tunnel progressed as well. The huge steel bulkheads, each about the height of a threestory building, were removed from between the tubes. First the steel plate and supporting beams were cut with oxyacetylene torches into sections small enough to be hauled out by a flatbed truck. Then, with two facing bulkheads removed at a joint and the closure plate in place and tested against leakage, concrete could be installed to span the joint area. In this way the tubes progressively became part of the continuous tunnel. After making a survey, engineers averaged out any small variations in alignment, and the curbs and sidewalks were constructed to this “fudged” line. This procedure gives the appearance of a smooth line even though the tunnel is built from prefabricated sections that may not always match perfectly at the joints.
A LOT OF WORK REMAINED TO BE done after all the tubes were joined together. The interior ceiling had to be erected, tile finish had to be installed, and all the other items of a complete tunnel had to be put in place and tested. Meanwhile, on land, the approaches, ventilation buildings (one of them a simple steel structure, the other faced with brick to be visually compatible with the fort), toll plaza, and administration buildings were built, and all the mechanical and electrical items were tested. Opening day was in November 1985.
Through the years the immersedtunnel method has been applied to some unique configurations. For instance, the Sixty-third Street Tunnel, across the East River in New York City, is square in cross section and carries two subway tunnels and two Long Island Railroad tunnels. The Chesapeake Bay Bridge and Tunnel in Virginia, which opened in 1964, is even more remarkable. This eclectic structure takes motorists from Virginia Beach to the Eastern Shore by means of two tunnels, two bridges, four islands, and fifteen miles of trestles. The ocean crossing would have been impossible to build using any other tunneling method.
In the Netherlands, where much of the land is below sea level, the immersed method is used even for tunnels that do not pass under a body of water. A section of the Rotterdam Metro was constructed by excavating a narrow canal down the center of town, into which immersed tubes were floated. These tubes were then sunk, joined, and covered, and the streets were restored. Recently at Amsterdam’s Schiphol Airport, the same method was used to install a tunnel under a main runway that could be out of service for only a short time. The tubes were fabricated in an adjacent excavation; then, during a season of low air traffic, a trench was cut through the runway and flooded. The tubes were quickly floated into place and covered, and the runway was rebuilt. This type of tunnel has also been used for utilities, such as pipelines, conveyor belts, and electric power.
MANY WAR STORIES EXIST FROM immersed-tunnel projects where Murphy’s Law has been obeyed. Tubes have been flooded and sunk; storms have sent tubes adrift; once a tube was made the wrong length; another time a tube broke loose from the lowering falls. A dredge cut through a four-footdiameter water main at the bottom of the harbor and came close to putting a city out of business. A tube slid down the trench in a storm and could not be put back into proper position, so an adjacent one had to be shortened. Despite all these troubles, the safety record of immersed tunneling has been excellent. During the five-year, $750million Fort McHenry Tunnel project, there were only two fatalities. Other tunneling methods have the sorts of risks normally associated with the mining industry.
If I had to pick the part of a construction organization most responsible for the success of an immersedtunnel project, it would be the survey crew. Everything depends on its work, much of which is done long before each tube is put down. For instance, the mating tube faces, typically forty feet high by eighty feet wide, have to be manufactured to very tight tolerances, and the surveyors check to see that they stay within a quarterinch of a true plane. This work must be done at night, after the ambient temperature stabilizes, since sunshine on one side of a tube will cause it to bend out of shape. The interior construction requires constant survey control as well.
In recent years a number of studies have been made for submerged floating tunnels—the somewhat startling idea of a floating tunnel tethered to the bottom in very deep water. The concept is particularly well suited to the quiet waters of deep fjords. In fact, there is a project under way in Norway for just such a tunnel across the Hogsfjord. Another spectacular project, still in the conceptual design stage, is a proposed crossing between Sicily and the boot of Italy. This would involve a system of three submerged floating tunnels across the Strait of Messina—two for cars plus a twotrack railroad tunnel. Neither of these projects is into construction yet, but where very deep water must be crossed and the span is too great for a bridge, the idea is very promising. It may take a little convincing, however, for the driving public to feel safe in a tunnel surrounded by water on all sides.
THE IMMERSED-TUNNEL METHOD continues to gain in popularity among civil engineers, and for many reasons. Compared with bored tunnels, it is safer, usually requires less total length, and allows more regular, predictable scheduling. Often it is the only feasible alternative where a bridge is not an option and navigation cannot be interrupted by a cofferdam to build a tunnel “in the dry.” Many world-class projects of this kind have been built in Europe and the Far East. Current examples under construction include the Western Harbor Crossing in Hong Kong and a forthcoming two-mile section of the Øresund Link between Denmark and Sweden. Yet for all its worldwide popularity, the technology was born and developed in the United States, and we still hold the record for the longest immersed tunnel in the world (in the BART system) and the largest (Fort McHenry).