Something about models fascinates people. I’m not talking about the “super” kind that strut on catwalks. I’m referring to the three-dimensional representation variety. Models are a staple of the heritage and museum industry – I always liked the dike and abattoir Acadian landscape model at the predecessor to the Canadian Museum of History, or the Plains of Abraham model at the Canadian War Museum. But models aren’t just used to represent history; they are also used to make history.
In the 1950s Canada and the United States cooperated to build two transborder megaprojects: the St. Lawrence Seaway and Power Project, and the Niagara Falls remedial works. I’ve written about both of them for The Otter~La loutre multiple times in the past, so I’m not going to rehash the details here. Instead, I’d like to focus on the high-precision scale models used in the construction of both of these massive water control endeavors since I think they are, for lack of a more sophisticated academic term, pretty cool.
The ideal way to conceptually approach these models and megaprojects is through the lens of envirotech, which combines the fields of the history of technology and environmental history. Envirotechnical analysis reveals the ways that models were integral to the creation of hybrid enviro-technical systems that blend the natural and the artificial, highlighting the engineering and modeling processes as a means of talking about conceptions of modernity, progress, technology, and environments in early Cold War North America.
The Canada-US Niagara River Diversion Treaty of 1950 called for remedial works and physical reconfigurations – what I like to call “disguised design” – that would maintain the beauty of Niagara Falls while allowing ½ to ¾ of the water of the Niagara River to be diverted around the falls for hydro-electric production, depending on the time of day. Scale models were the primary means by which the form and location of the Niagara works were chosen. The Niagara models replicated long stretches of the river and falls in minute detail: the topography, the shoreline, the channels and contours, the cataracts and rapids, and the turbulence and velocity of the currents. This appears to the be the first time in Canada that models were used as the basis for a civil engineering project, though I’d like to get some more definitive confirmation of that.
Two models were used for Niagara planning, one by each country. The American model was built by the US Army Corps of Engineers at the Waterways Experiment Station at Vicksburg, Mississippi. This station was built in 1928 to model the Mississippi River (and a recent piece on the Mississippi River models partially inspired this post). Canada’s Niagara model was built by Ontario Hydro at its Islington station, on Kipling Avenue in Toronto. The Ontario model depicted 5 miles of the river, from the tip of Grand Island to Rainbow Bridge, including the falls. Encased within concrete walls, it was 95 feet long, 37 feet wide, and approximately 4 feet high, built on a horizontal scale of 250:1 and a 50:1 vertical scale. The Vicksburg model included the upper reaches of the Niagara River, and since it included more territory it was on a smaller scale (360:1 horizontal; 60:1 vertical) but was 260 feet long.
Planners were enamored with the models and believed them to be indispensable for determining the future fluvial geomorphology of the two iconic waterscapes. Moreover, it was cost-effective: Ontario Hydro bragged that using models saved about $5 million. Both nations contended that the use of separate models increased accuracy due to the “uniqueness and complexity” of the situation. But politics and pride were also motivations for using separate models: each country wanted a piece of the action, and thought their experts and means were more reliable.
While construction of the Niagara remedial works, based on the models, was underway, an agreement for the transnational construction of the St. Lawrence Seaway and Power Project was finally inked. Most of the governmental agencies responsible for Niagara were also involved with the St. Lawrence, and many of the Niagara engineers were moved to the St. Lawrence models. Ontario Hydro built 9 power project models, again at Islington. The federal St. Lawrence Seaway Authority built its own models for the seaway, in both Ottawa and Montreal. The National Research Council (NRC) in Ottawa created a model of a lock and another of the International Rapids section. The United States undertook hydraulic model studies in Vicksburg, and also in Minneapolis, Minnesota.
Modelling a longer stretch of the St. Lawrence system, and one that was to be converted into a lake-cum-reservoir, provided modeling challenges. As had been the case with Niagara, part of the problem was finding accurate information on which to base the construction of models. The engineers spent a great deal of time trying to find out what the “natural conditions” were, but internal documents reveal top officials admitting that they didn’t really know. Technocrats qualified their control of water levels with the vague phrase “as nearly as may be”; however, this imprecision was belied by public displays of confidence and authority.
The engineering process was beset by other miscalculations, assumptions, and compromises. The St. Lawrence and Niagara models often produced erroneous results. Sometimes this was the result of incorrect base knowledge, such as faulty gauge data, about the rivers. This hampered the “model verification” since that required comparing model performance against known conditions. While the engineers came to personally know the river environments in great detail, embodied and experiential local knowledge was ignored and dismissed. Technical problems were also exacerbated by political rivalries: there were many instances of Canadian engineers complaining about the inadequacy of American tests and models, and vice versa.
When extrapolated onto a larger scale, seemingly slight model mistakes could have significant ramifications. An internal Ontario Hydro engineering report concluded that: “on the basis of Niagara tests which Hydro itself has run on models of its own design, Hydro Engineers have been led to conclude that the models being used by the Board, which were designed to a distorted scale, produce results which are quite different from those run on true scale models.” During construction of the Seaway “it was discovered that, due to an oversight in establishing … conditions, an area comprising approximately 1½ blocks was left in the lower channel.”
The engineering plans and models for the two megaprojects were indicative of a faith in progress and technology, exemplified by the models the planners relied upon; but the evolution of these plans equally reveal the contradictions and limits to this faith. Even though the engineers were repeatedly made aware of their fallibility and the limits of their expertise, they continued to ignore the underlying flaws in their methods and maintained full trust in their models and technological expertise. At the same time, though the engineering process was messy and reactive, we shouldn’t forget that the projects did function as the engineers had planned in the long run.
I think it is fair to describe this modeling process as indicative of high modernism. The concept’s originator, James C. Scott, argues that truly (or “ultra”) high modernist projects can really only happen where authoritarian governments rule over weak civil societies. My point of departure is that the St. Lawrence and Niagara projects do meet the major criteria of high modernist projects, aside of the authoritarian setting, and aren’t thwarted, though they are modified, by the exigencies of liberal capitalist democracies. Thus we see what I have previously called negotiated high modernism: lacking the autocratic authority to simply impose schemes without some measure of approval from civil society, the Canadian (and American) states had to repeatedly negotiate, adapt, and legitimize themselves – in relation to both the specific natural environments and the societies they aimed to control – in order to achieve their high modernist vision.
Though many of the models were made available to the public during the 1950s, and proved to be a very popular draw. While the fate of most of the Niagara and St. Lawrence models is unclear, the Corps of Engineers kept their Niagara model at the Waterways Experiment Station in Vicksburg intact and open to the public into the 21st century. Despite my pestering, I haven’t received any confirmation that the Niagara models are still in existence. There are rumours that the National Research Council’s St. Lawrence models might still be around, disassembled and boxed up in some government warehouse; government historians have been looking into the case for me, but so far no definitive proof has surfaced. Until they turn up, we will have to be satisfied with viewing the real deal – even if the Niagara and St. Lawrence waterscapes are, in certain respects, as much technological artifacts as the models.
Latest posts by Daniel Macfarlane (see all)
- Canadian Environmental History at ASEH 2023 - March 21, 2023
- Environmental Studies of the Great Lakes: A Symposium - October 27, 2022
- The Great Lakes Water Quality Agreement Turns 50 - April 14, 2022
- New Book Series: Environmental Studies of the Great Lakes - March 1, 2022
- Graduate Student Opportunity with the Canadian Museum of Nature - February 2, 2022
- The Climate Crisis and the Canadian Classroom - September 21, 2021
- Line 5 and the History of Canada-US Hydrocarbon Pipelines - June 17, 2021
- Hoover Dam in Hollywood - December 3, 2020
- Environmental Diplomacy, Envirotech, and Canadian History - May 5, 2020
- Canadians at ASEH & N|H|S Get Together - March 7, 2020
great post Dan — these large-scale models are very cool. In my reading about models, I have not encountered ones of this size or complexity. It would be worthwhile to investigate the history of computer modelling and track how that form of modelling came to replace this Lilliputian physical approach! And I too would love to know where they are! Imagine putting them on exhibition…
Great article. The NRC model of the St. Lawrence River was, like most physical models, used as part of an engineering study. Once completed, the models were disassembled and relegated to the proverbial dustbin while the space in the laboratory was re-purposed to the next challenge. I believe a physical model of the lower Fraser River followed the St. Lawrence River study. Over time these large models were replaced with computer models. For a while ‘hybrid models’ were used wherein the larger river domain was modeled using a computer model which controlled the boundary conditions (flows and water levels) for a detailed physical model of the key area of study. At Vicksburg, it is more common to build a model of a physical space (St. Lawrence River, Port of Los Angeles, etc.) and to keep it in perpetuity – revisiting the model decade after decade as new challenges or questions arise. That’s in part a luxury that Vicksburg can enjoy courtesy of its mild climate and use of outdoor facilities. Up in Ottawa, the weather drives us toward indoor models where space is at a premium.
There are a great many interesting stories about the building of the Seaway and of the technical back-and-forth between the American and Canadian technical specialists in their struggle to share the development equitably.
W.H. Becker’s “A History of the US Army Corps of Engineers and the St. Lawrence Seaway” (1984) provides some interesting insights. The history of the National Research Council Hydraulics Lab (where I worked as a researcher for many years) is intricately tied to the Seaway. Work on which formed the basis for decades of innovative hydraulic research.
Physical models are still used for rivers and coasts, mainly their role now is for complex flow problems and as research tools for development and validation of improved computational models. Essentially, they are a stepping stone in the development of technology. Observation, experimentation (including physical modeling), then development of theories (and computer models)… River engineering has largely matured to the state that computational models can reliably used now to solve problems that a few decades ago would have required massive physical models. The focus of physical modelling is now on complex unsteady flow patterns, breaking waves, ice forces, etc. That said, in those cases where a physical model can be justified due to project complexities and uncertainties – it is enormously satisfying to build a scale model of a little portion of the world and to carefully calibrate flows and levels such that they reproduce reality. And then, of course, the main part – to imagine a new reality (a realigned channel, a new harbour, etc.) and to use the model to explore how it responds to waves, currents, storms, etc.
Thanks for your comments Will and Mike! And thanks very much for your insight, Mike. I build on Becker in my own book on the creation of the Seaway (the parts about the Seaway in this post are largely drawn from that book), and Marty Reuss has done some great work on USACE models. Feel free to drop me a line if you are interested, as I’m finishing up a book on the engineering of Niagara Falls and would love to hear more about your experiences and knowledge of the evolution of hydraulic modeling.
Dear Mr. Macfarlane, Sadly, the Niagara Falls model at the Waterways Experiment Station (WES) in Vicksburg was dismantled. In the aftermath of the terrorist attack on the World Trade Center in 2001, WES was closed to the public and the Niagara Falls model and the building in which it was located were dismantled.
Terry Winschel, Historian, U.S. Army Corps of Engineers Research and Development Center, Vicksburg, MS
I’m sad to hear that, Terry, but thanks so much for updating me on this.