In late summer 1928, a worker on a construction barge moored in the middle of the Detroit River looked at the surface of the water and could not see the bottom. He did not need to see it. The bottom lay twenty-eight feet below gray-green water that moved in a rhythm he could watch even from the barge’s railing, flowing southwest toward Lake Erie, just as it had flowed since the last glaciers cut this channel twelve thousand years before. The current was not fast in the way that floods are fast.
It was constant. It was the drainage of four of the five Great Lakes—Superior, Michigan, Huron, and Erie—passing through a channel a quarter mile wide between the city of Detroit, Michigan, on the north bank and the city of Windsor, Ontario, on the south bank, through what was in 1928 one of the busiest commercial waterways on earth. It was a channel that carried more shipping by tonnage in some years than the Suez Canal, and it was not about to stop running while a construction crew tried to build a tunnel beneath it. The man on the barge was part of a dredging crew preparing the riverbed to receive sections of steel tube that would be lowered into place in the coming weeks.
The dredge’s bucket went down into the current and came back up loaded with material from the river floor. The riverbed was fine clay and silt deposited over centuries of slow accumulation from the drainage of the Great Lakes basin. The clay and silt had to be removed before the steel tube could be lowered. The steel tube had to be lowered before there could be a tunnel.
The tunnel was the reason the barge sat in the middle of the Detroit River that summer day in 1928, with ore freighters passing close enough on their way to the Soo Locks that the dredging crew could read the names painted on their hulls. The barge was positioned roughly sixty feet above where the centerline of the tunnel would eventually pass. The tunnel did not yet exist. The trench that would receive it did not exist either.
The steel tube sections that would eventually fill that trench were being built at that very moment in a yard on the American shore, fabricated from structural steel and reinforced concrete, assembled in lengths up to 260 feet. Each section was wide enough inside to carry two lanes of vehicle traffic side by side. Within about fourteen months, one of those tube sections would be floated into position directly above where this man stood, held in place by cables and winches and the men who controlled them against a current that never stopped, then sunk into the trench the dredge would have cut into the riverbed clay by then. Within about twenty-six months, the tunnel of which that tube section would become a part would open to vehicle traffic between the United States and Canada.
The man on the barge would be somewhere else. His name, his wages, and his hours on this project would exist in the employment records of the contracting company that paid him, in whatever form those records survived. This is the story of how two cities in two countries separated by one river built the world’s first international underwater vehicle tunnel in twenty-six months, across a waterway that never stopped running, and through a riverbed that provided no stable foundation without being engineered and designed. It is the story of the immersed tube, a method of crossing a body of water that required neither pumping the water out nor stopping the commercial traffic on it, but instead required building the tunnel on shore, floating it into position in the river, and lowering it through twenty-eight feet of moving water into a trench cut from the bottom of one of the busiest commercial channels on earth, with a precision that allowed adjacent sections to meet underwater within inches of their designated positions.
It is the story of the men who worked on the dredges in the current, the men who dug the approach tunnels through the clay beneath both cities in compressed air, and the men who maneuvered the tube sections in the river while the current pulled against the cables holding them in place. The tunnel they built carries vehicle traffic under the Detroit River today. Most of the people driving through it have never thought about the men who built it. That is how infrastructure works.
To understand why two cities in two countries needed a tunnel under the river between them, you have to understand what the Detroit River was and what crossing it cost before the tunnel existed. Detroit sits on the northwest bank of the Detroit River, on land that rises gently from the water’s edge inland across southeastern Michigan. Windsor, Ontario, sits directly across the river on the southeast bank, on the flat agricultural plain of Essex County, at the southernmost tip of Canada. The geography is counterintuitive to anyone who has not studied a map carefully.
Windsor is in Canada. Windsor is also directly south of Detroit, which is in the United States. The border between the two countries runs through the middle of the river. The river flows almost due southwest from Lake St.
Clair toward Lake Erie, carrying the waters of the upper Great Lakes basin through a channel averaging a quarter mile wide and twenty-eight feet deep. In 1928, the two cities on either side of that channel were among the most economically interconnected in North America. Detroit was the center of American automobile manufacturing. Ford had built the River Rouge complex on the American shore.
General Motors’ headquarters was in downtown Detroit. Chrysler had consolidated its operations in the metropolitan area. The Canadian automobile industry, centered in Windsor and the surrounding region, worked in close commercial relationship with the American factories across the river. Parts manufactured in Windsor were assembled in Detroit.
Subassemblies manufactured in Detroit were finished in Windsor. Workers who lived on one side of the border worked on the other. The river between them was an international boundary, and for the purposes of the daily commerce of the two-nation industrial economy surrounding it, a boundary crossed constantly in both directions by vehicles carrying goods and people whose livelihoods depended on the crossing being as fast and reliable as the industrial process on both sides demanded. Crossing the river in 1928 meant one of two things.
It meant the ferry. The ferries ran on schedules and carried limited numbers of vehicles, were subject to the river’s ice conditions in winter, and at the peak of the autumn shipping season, when industrial traffic between the two cities was at its densest, vehicle queues built up at both docks with waiting times measured in hours. Commercial vehicle operators, paying drivers by the hour to deliver time-sensitive goods to factories across the river, did not consider multi-hour ferry waits an acceptable operating cost. Or it meant the Ambassador Bridge, which was under construction at that very moment downstream from the tunnel site and would open in November 1929, a year after the tunnel construction began.
The bridge would provide a fixed vehicle connection over the river for the first time. A suspension bridge over a major commercial waterway required a deck clearance dictated by the largest ships in the commercial traffic. The Ambassador Bridge was designed for lake ships drawing up to twenty-five feet. The bridge deck hung high above the river.
The approach ramps to a deck that high consumed land in both cities and imposed a vertical climb on every heavy vehicle that crossed it. In January, when temperatures fell below freezing and the approach ramps iced over, the bridge was a high fixed crossing exposed to everything a continental winter produces over an open body of water. A tunnel under the Detroit River in January is a tunnel. The conditions inside do not change because the weather outside changes.
The first formal proposal for a fixed crossing under the Detroit River appeared in the 1870s, when the techniques that had made the St. Clair River railroad tunnel approach theoretically possible were being applied in New York City to the approaches of the Brooklyn Bridge. Engineers discussed the possibility. Those discussions did not lead to the construction of anything.
The traffic volumes of the 1870s did not generate the commercial justification for the financing such an underwater crossing required, and the technique for the specific challenge of the Detroit River—a wide, fast, commercially active waterway with a soft clay bottom, too deep for simple temporary cofferdams—had not yet been organized into a method the engineering profession could price with confidence. Discussions resumed in the 1890s. They were renewed again in the first decade of the twentieth century, when the Michigan Central Railroad built a railway tunnel under the Detroit River using the immersed tube method and opened it in 1910. The Michigan Central tunnel proved the method worked under Detroit River conditions.
But it did not prove the method could be applied to a vehicle tunnel of the size required by 1920s automobile traffic, within the financial and regulatory framework of a project crossing an international boundary and requiring agreement between two sovereign governments. The international border was not a simple complication. A tunnel under the Detroit River was a structure that began in one country, passed under the boundary separating the two countries, and ended in another. The legal authority to build and operate that structure and collect tolls from it required an agreement between the governments of the United States and Canada at a level above the municipalities of Detroit and Windsor, and above the state government of Michigan and the provincial government of Ontario.
It required a treaty or its equivalent. The diplomatic and legal apparatus of two national governments had to engage with what was, at the level of its construction, a municipal infrastructure project. That engagement took time. It took until 1920 for the legal framework of the joint project to be established by an act of the United States Congress and a corresponding act of the Canadian Parliament.
It took until 1928 for the financing mechanism to be assembled and construction to begin. What changed by 1928 was traffic and the willingness of the bond market to finance a project against it. The automobile had reached the mass market two decades earlier. By 1928, traffic across the Detroit River by ferry had grown to a volume the ferry fleet was clearly unable to handle.
Toll tunnel revenue projections based on the traffic between Detroit and Windsor could be presented to buyers with a confidence that the pre-automobile projections of earlier decades could not support. Bond buyers bought the securities. The city of Detroit and the city of Windsor each agreed to their shares of the financing. Construction began.
The engineering challenge of crossing the Detroit River by tunnel had two distinct components that met at a point below each bank of the river. The river crossing itself, the portion of the tunnel that would run under the water, was built by the immersed tube method. The approach sections on each side, running under the streets of Detroit on the American side and under the streets of Windsor on the Canadian side, were excavated through the clay beneath each city using a compressed-air tunnel shield. The logic of the immersed tube method was straightforward.
If you had to place a tube across a body of water, and the body of water was too disturbed to work in directly from the surface, you built the tube on dry land, sealed its ends, floated it to where it needed to go, positioned it precisely over a trench you had dredged into the riverbed in advance, sank it into the trench by flooding the interior of the tube with controlled water, connected it to the neighboring sections already in the trench, and then pumped the remaining water out. Then the tube was on the riverbed, sealed, and connected from both ends, and you had a dry crossing under the river without having worked in the river in any direct way. The Michigan Central had proven the method in 1910 with a railway tunnel at the same crossing. The 1928 project applied the method at a larger scale for vehicles.
The tube sections for the vehicle tunnel were built in a fabrication yard established on the American shore downstream from the crossing route. Each section was a structural steel cylinder with an interior diameter of twenty-eight feet, large enough for two lanes of vehicle traffic at once, encased in reinforced concrete providing structural mass and chemical resistance to the river water. The sections were built in lengths of approximately 260 feet. Each section, when complete, was sealed at both ends with temporary steel bulkheads trapping air inside and keeping the section afloat.
A section 260 feet long and 28 feet in diameter, sealed and filled with air, floats. It sits low in the water because of its weight, but it floats. It could be moved by tugs. The dredge worked the riverbed ahead of each tube placement, cutting the trench through the clay at the depth and grade the engineers had specified.
The Detroit River bottom clay was soft enough for the dredge bucket to cut through without the difficulties of harder material. The clay was also cohesive enough to hold the shape of the trench walls during the interval between dredging and tube placement. The current flowed through the open trench continuously and redeposited material into it as it ran. The dredge made additional passes to remove what the current had redeposited.
The relationship between the dredge and the current was continuous throughout the construction of the river section. The current did not stop to make way for the dredge. The dredge worked accommodating the current. When a tube section was ready, it was floated from the fabrication yard at a time chosen for the current and tide.
The Detroit River carries a slight tidal influence from Lake Erie, and the state of the tide affected the ease of floating and towing the huge objects by tugs. The section had to be placed over the trench with accuracy measured in inches. A current of two to four miles per hour against a body 260 feet long and 28 feet in diameter generated a lateral force the tugs and positioning cables had to resist continuously. Mooring cables ran from the tube section to anchors on both banks, and cable tension was adjusted as the section was guided over the trench.
When position was confirmed by survey, with surveyors on both banks taking synchronized readings and reporting to the positioning crew, the controlled sinking began. Water was admitted through valves at specified rates while air was vented through other valves. As buoyancy decreased and the interior filled, the section settled toward the trench. The men in charge of the valves worked from the top of the section, which remained above the river surface through most of the early sinking stages.
The moment the section’s deck disappeared beneath the water, the positioning crew above it lost their direct visual reference. Below the surface, positioning continued through instruments, cable tension, surveyor readings from the banks, and the work of divers who descended to watch the sections descend and approach the trench bottom. The divers worked in the Detroit River current at depths of twenty to thirty feet, in limited visibility due to the turbid, silt-laden nature of the waterway. They worked by touch and by signals they sent to the surface.
When the section settled into the trench on prepared bearing foundations, the divers connected it to the previously placed section. The connection required the temporary bulkhead at the leading end of the new section to be brought into contact with the corresponding bulkhead at the trailing end of the section already in the trench. A gasket system allowed the two faces to seal tight against each other. Once the seal was confirmed, water was pumped from the interior of the new section through a series of pumps and valves.
The differential in water pressure on the face of the joint helped compress the gasket as the interior pressure equalized with atmospheric pressure. When the interior was dry, workers could enter through the previously placed section, walk to the joint, and pour the permanent concrete that would fix the connection for the life of the structure. This sequence—dredging, fabrication, floating, positioning, sinking, connecting, dewatering, and finishing—was the sequence followed for each section of the river crossing. Sections were lowered from both shores simultaneously, with the section coming from the American side and the section coming from the Canadian side meeting in the middle of the river.
That meeting point was the most precise moment in the entire construction sequence. The final section had to fit between the two sections approaching from each side within the tolerances the engineers had set. It fit. The current moved over the completed sections as they were covered with backfill and protective stone.
The ships passing through the commercial channel over the construction site, which had deviated slightly from their usual course during active work phases, returned to their normal path. The river did not stop during construction. The work was done around the river. The approach tunnels beneath the two cities were built by different means and under different conditions.
The ground beneath Detroit and beneath Windsor at the depth of the tunnel route is glacial lake clay, deposited when the Great Lakes basins were covered by much larger bodies of water in the millennia following the last ice age. This clay, when properly supported, is dense and cohesive and will hold the shape of a tunnel cavity if lined promptly after excavation, but it contains water. The hydrostatic pressure of the groundwater in the clay at tunnel depth, combined with the hydraulic connection between the groundwater and the river twenty to thirty feet above, meant that any opening made in the clay would receive water through the clay matrix unless the driving pressure of the water was resisted by equal or greater pressure in the work space. That counter-pressure was compressed air.
Workers in the approach tunnel sections worked in air at pressures above atmospheric, raised to a level sufficient to resist the hydrostatic pressure of the groundwater in the surrounding clay. The tunnel shield, a cylindrical steel device with a cutting edge at its face and a hydraulic system that pushed it progressively through the clay, excavated the material ahead of the lined sections behind it. Workers inside the shield removed the loosened clay, loaded it into small cars that traveled back through the compressed section to an airlock at the unpressurized end of the tunnel. The airlock was the pressure boundary.
Workers and materials moved between the pressurized working face and the completed unpressurized tunnel through the lock, spending the necessary time at each pressure stage before moving to the next. Working in compressed air at the pressures required in the approach sections of the Detroit-Windsor tunnel meant that every man who entered the working face was exposed to the physiological hazard that had killed and permanently disabled workers in bridge caissons in the previous century. Decompression sickness, the bends, is nitrogen dissolved in the blood and body tissues at elevated pressure that forms bubbles on tissue walls if the pressure drops too rapidly during exit through the lock. The worker who decompressed too quickly, who moved through the pressure stages of the exit sequence faster than the schedule specified, did not give the nitrogen enough time to diffuse across the lung tissues before it came out of solution as bubbles in the joints, or in the spinal cord, or in the arterial supply to the brain.
A man bent in the joints recovered with pain and sometimes with permanent damage to the joint tissue. A man bent in the spinal cord could lose function of the limbs below the affected level. A man bent in the cerebral circulation could die. Medical understanding of decompression sickness in 1928 was far more advanced than it had been during the construction of the Brooklyn Bridge caissons in the 1870s, when workers died of the bends in large numbers because the disease was not understood and the mechanism of injury was unknown.
By 1928, decompression schedules specifying minimum exit times at each pressure stage had been developed through British Royal Navy research in the early decades of the twentieth century, and refined by American practitioners in the tunnel construction industry through years of experience in underwater construction on port projects and railway tunnels since the 1890s. The schedule set the minimum decompression time for each combination of working pressure and shift length. If the schedules were followed correctly, the risk was reduced significantly. If the air pressure in the working section was maintained accurately at the specified level.
If the individual worker’s physiology fell within the subset that did not respond badly to dissolved nitrogen, no matter how precisely the protocol was followed. The men who worked in the compressed-air sections of the Detroit and Windsor approach tunnels were paid a compressed-air premium above the wages of surface and free-air workers. A wage differential calculated in 1928 to cover the additional risks they agreed to accept in exchange for doing this work. Whether that differential adequately compensated for those risks is a question the payroll records do not answer.
The payroll records document the rate only. Either the men accepted the rate or they looked for other work. The workers who accepted the compressed-air work passed through the airlock at the start of each shift and exited through it at the end, spending the decompression stage times required by the schedules before returning to atmospheric pressure and coming back up to the surface. Detroit’s winter complicated every aspect of the outdoor river work.
The Detroit River does not freeze solid because the volume and speed of the water passing through the channel is too great, but it carries ice, ice from the surface of Lake St. Clair, which lies upstream and is shallower and slower and freezes in the bitter January cold of a Michigan winter, the ice breaking up and moving into the Detroit River channel and heading southwest with the current toward Lake Erie. This ice collected around the mooring cables of the positioning system. It piled against the hulls of barges and tugs.
It accumulated along the construction infrastructure near both shores where the current slowed. The tugs that had held the positioning system against the current in open summer water held it against the current and against the ice in winter. The dredge that cut the trench in summer worked amid the ice in winter. The survey crew that maintained positioning control from both banks operated their instruments from platforms that required constant snow and ice removal to function.
The men who worked the tube-placing operations from barge decks in December and January worked in continental winter conditions over an open body of water, with no shelter from the wind coming off Lake Erie to the southwest. In summer, the river was different. The summer of 1929 in the Great Lakes region was a summer of extreme heat and humidity over the water. Barge decks absorbed the heat of the midday sun and held it into the evening.
The men who worked in the fabrication yard on the American shore worked in a steel-and-concrete yard under direct sun, with the surfaces of the tube sections absorbing heat all day, too hot to touch without protection in the afternoon. There was no artificial shade on a fabrication barge while working. There was water in every direction, and there was the heat the water itself generated under a Michigan summer sun. The workforce that built the Detroit-Windsor tunnel came from the industrial labor pool of the greater Detroit area, which in 1928 was one of the largest and most diverse manufacturing labor markets in the world.
Detroit in the 1920s was a city transformed by two decades of explosive automobile industry growth. Men arrived from southern and eastern Europe, Poles, Hungarians, Italians, Czechs, and Romanians, reaching the city in the hundreds of thousands between 1900 and the immigration restrictions of 1924, driven by the automobile factory wages paid by Ford, General Motors, and the supplier network. They settled in the neighborhoods of the east and west sides, built communities, and raised children who were now entering the labor market themselves. The construction labor pool drew continuously from these communities.
A worker on a Ford foundry line one year might move to a construction crew the next, following wages wherever the work paid. Black workers from the American South had been arriving in Detroit since the early years of the Great Migration, the movement of Black Americans out of the rural South that had run since approximately 1915 and accelerated through the 1920s as economic conditions under Jim Crow pushed and industrial wages in northern cities pulled. Ford employed Black workers at River Rouge at higher rates than most major American industrial employers of that era, though typically in the most physically demanding, heat-intensive departments such as the foundry, coke ovens, and rolling mill, rather than the assembly trades where the occupational classification rewards concentrated. The Great Migration created large Black communities in Detroit’s east-side neighborhoods, communities whose residents needed work, and whose access to skilled trades was restricted by the union admission practices of that era.
The Detroit construction industry in 1928 reflected the racial organization of the broader American labor market. The skilled trade jobs—the ironworkers who fabricated the tube sections, the licensed equipment operators who ran the cranes and dredges, the carpenters who built the forms, the licensed electricians—were held predominantly by white workers, many of them members of craft unions whose admission policies in 1928 were designed to exclude Black applicants from membership, and therefore from the trades. The common labor jobs—moving materials, removing excavated clay from the tunnel shield, performing the physical tasks construction requires without the classification of a recognized trade—were done by Black workers and new immigrants from southern and eastern Europe. Their wages reflected their place in the occupational structure, higher than the agricultural and domestic service workers many of them had come from, lower than the craftsmen above them.
The tunnel shield in the compressed-air section did not distinguish between men across the racial line. The air pressure was the same throughout the working face. The decompression schedules applied to everyone who exited through the airlock. Whether the likelihood of a dangerous decompression was equal among all workers, whether the enforcement of the schedules was equal, and whether the medical supervision the compressed-air premium was supposed to reflect actually attended to the common laborer as closely as it did to the skilled operator, is a question the construction accounts in any widely available public record do not answer.
The payroll records of the contracting companies that built the tunnel contain the names of the men who worked on the river, the names of the men who ran the dredge in the current of the winter of 1928, the names of the men who handled the tube-positioning cables in the river as the sections were lowered into place, the names of the men who entered the airlock at the start of each shift and exited at the end, and the names of the men who did not exit in the same condition they entered. The deaths and injuries associated with the construction of the Detroit-Windsor tunnel are not comprehensively documented in any widely circulated public report of the project. Compressed-air work in the late 1920s was a known hazard with a known protocol and a known failure rate even when the protocol was followed correctly. The regulatory environment governing construction worker safety in Michigan and Ontario in 1928 required the reporting of serious injuries and fatalities.
That environment did not require that those records be kept in a form easily accessible to researchers ninety-five years later. The employment practices of the late-1920s construction industry did not generate the kind of continuous safety documentation that later regulatory frameworks would require. What the records show is that the work was done, that the trench was dredged, that the tube sections were placed in it, that the approach tunnels reached the river sections and connected to them. The structure was completed to the required specifications.
The cost in lives of the men who built it is another account, and that different account is not answered by the records that have been made available. On November 3, 1930, the Detroit-Windsor tunnel opened to vehicle traffic. Twenty-six months after the dredge made its first pass across the riverbed, and the construction barge crew went out to work on a river they could not see through, the tunnel beneath it was complete and open. The opening ceremony was held at both portals simultaneously.
On the American side, at the Jefferson Avenue entrance in Detroit, officials of the city of Detroit and the state of Michigan gathered for photographs. On the Canadian side, at the Wyandotte Street entrance in Windsor, officials of the city of Windsor and the province of Ontario gathered likewise. The mayors of both cities were present. The figures of the joint financing authority that had organized the bond structure were present.
The engineers whose designs were now realized in concrete and steel and in the clay of the riverbed beneath the assembled officials’ feet were present. Photographers recorded the moment from both sides of the border. Wire services carried the story. Both cities ran the opening on their front pages.
The first vehicle crossed the moment officials at both ends declared the tunnel open. The tunnel was 5,160 feet long, just short of a mile, with the immersed tube section running approximately 2,200 feet beneath the flowing waters of the Detroit River at a maximum depth of seventy-five feet below the river surface. Crossing that distance took about three minutes at the speed limit. The structure that the tube crews had placed in the river’s current, that the dredging crews had cut the trench for through the clay beneath two cities, and that the finishing crews had completed with concrete, tile, and lighting, was now, in the moment of passage, a road.
A road under a river. A road under an international border. A road between two countries connected by a mile of tube in the bed of a river that had separated them commercially and geographically since the first European settlement on both banks. The first international vehicle tunnel had been built.
No nation before 1930 had built a fixed vehicle crossing under a river forming an international border. The Detroit River had formed the boundary between the United States and Canada since 1783. That boundary had been crossed by boat since before the American republic or the Canadian confederation existed as political entities. On November 3, 1930, it was crossed through a tunnel consisting of a mile of reinforced concrete tube in the riverbed clay, carrying two lanes of traffic in each direction, ventilated by forced-air systems designed to keep the carbon monoxide from the internal combustion engines of those vehicles below lethal concentration levels, lit by electric lamps, and operated by toll collectors at both ends.
The men who built it were not in Windsor or Detroit on November 3, 1930. A construction workforce does not gather at the opening ceremony of a twenty-six-month project to acknowledge its contribution. The construction workforce had appeared on the payrolls during the period of construction. When construction ended, the payroll ended.
The men who had run the dredge in the current were on the next dredging job, wherever that was. The men who had worked the tube-placing operations on the river were on the next marine project in the Great Lakes, or the inland waterways system, or the Gulf Coast ports. The men who had walked through the airlock into the compressed air of the approach tunnels were on the next compressed-air project, wherever it began. The men who had moved materials, removed excavation, and done the heavy work the project required were looking for the next job that would take them.
The engineers whose names appeared in the design records received the professional recognition that designing a major piece of infrastructure brings in the technical literature of the profession. The civic officials whose names appeared in the construction agreement and the bond contract received the recognition that elected and appointed officials receive in the administrative records of the cities they served. The mayors who presided over the opening ceremony appear in the municipal histories of Detroit and Windsor as officials present at the opening of the first fixed international vehicle crossing in this part of the continent. The men who worked the river sections in the ice of the winter of 1929 are in the payroll records of the dredging contractor and the tube-placing contractor.
The men who worked the approach tunnels in compressed air are in the payroll records of the shield-driving contractor. Their names are not on the commemorative plaques at either portal that bear the names of commissioners, bond trustees, and chief engineers. They do not appear in the record books that recognize the crossing as the first of its kind. They appear in the payroll records.
In whatever form those records survived the ninety-four years since construction completed. The Detroit-Windsor tunnel has operated continuously since November 3, 1930. It is ninety-four years old. The tube sections these men lowered into the current of the Detroit River in 1929 and early 1930 have been in the riverbed clay for ninety-four years.
The current of the Detroit River passes over them now as it passed over them when the dredge first cut the trench. At two to four miles per hour southwest toward Lake Erie, carrying the drainage of four of the Great Lakes. The water temperature over those tube sections approaches freezing in January and reaches the seventies Fahrenheit in August. River ice passes over them in winter.
Cargo-laden ships pass over them on their way to the Soo Locks. The chemistry of the river water, its dissolved mineral content, its seasonal salinity variations from the industrial and agricultural discharge of the surrounding watershed, has worked on the concrete outer surface of the tube sections for ninety-four years. Inspection crews assessing the structure on scheduled maintenance cycles examine the concrete surface, measure carbonation depth in the outer layers, record the condition of waterproofing membranes, note any cracking or spalling, and file reports with the maintenance authority deciding what needs repair and when. Sections that needed repair have been repaired.
The tunnel is still in service. The ventilation system has been upgraded from the 1930 specification. The fleets of vehicles crossing the tunnel today run internal combustion engines producing the same carbon monoxide they produced in 1930, but in quantities the original ventilation calculations did not anticipate. The ventilation buildings at each end have been modified.
The mechanical systems moving air through the tunnel have been upgraded. The toll collection system has been replaced several times. The structural condition monitoring technology the maintenance authority uses today did not exist when the tube sections were placed. The engineers who designed the 1930 systems could not have anticipated any of it.
The tube they placed those systems inside has not changed. The tube is what it was. You can drive through the Detroit-Windsor tunnel today. The crossing takes about three minutes in light traffic.
The entrance roadway at the Detroit portal descends from Jefferson Avenue at a grade that takes the road from street level down through the clay beneath the city and then beneath the river. At the lowest point of the crossing, the roadway lies approximately seventy-five feet below the river surface above. The riverbed lies twenty-eight feet below the river surface. The tube rests thirty-five feet below the riverbed in the bearing clay the dredge prepared for it in 1929.
Between the keel of the deepest ship using the commercial channel above and the roof of the tunnel below, there are approximately eighty feet of river and clay. From the middle of the river section, if you could stop, which the tunnel authority does not permit and the traffic behind you would not allow, neither portal would be visible. The tunnel curves slightly as it crosses to align with the river, and the lighting fixtures recede in both directions to the point where the bore closes in distance and light. The men who dug the approach tunnels through the clay of Detroit and Windsor could not see the portal they entered from or the portal they were working toward from the working face.
They could see the clay ahead of the shield’s cutting edge and the lined sections behind them and the string of work lights receding toward the airlock. Instruments told them where they were. Blow counts told them what the clay was doing. Decompression tables told them how long they had to spend in the airlock before they could exit back to the surface.
They worked toward a river they could not see, beneath a river they could not see, toward a city in another country they could not see. Survey markers and instruments gave them the correct heading. They followed the heading because following it accurately was the difference between the two approach drives meeting at the correct location within the river section, within the tolerances that allowed the joints to seal, and meeting at a location requiring either remediation or a permanent defect in the alignment of the finished structure. The two drives met within the specified tolerances.
The joints sealed. The tube sections the positioning crews lowered into the trench in the middle of the Detroit River current in 1929 rest in the riverbed clay. The approach tunnels excavated through the clay of both cities under compressed air are lined and functioning. The men who worked on the dredge and the positioning cables and the tunnel shields and the airlock sequences are in the payroll records of the companies that employed them, in whatever form those records survived.
The tunnel is 5,160 feet long. It was built in twenty-six months. It opened on November 3, 1930. It was the first vehicle tunnel to pass beneath an international border anywhere on earth.
The current of the Detroit River runs over it at two to four miles per hour. In January, ice comes down from Lake St. Clair and flows with the current over the tube sections buried in the clay below. In August, the water temperature over those sections approaches seventy degrees.
The chemistry of the river has worked on the concrete of those sections for ninety-four years. The tube sections are still in the clay.
