Transcona Grain Elevator — a Million Bushels That Sheared the Clay and Tipped 27 Degrees

Summary

On 18 October 1913, a brand-new million-bushel reinforced-concrete grain elevator at North Transcona, about 11 kilometres north-east of Winnipeg, sank bodily into the prairie and tilted nearly 27 degrees within a single day, killing no one but writing itself permanently into the foundations of soil mechanics. The 65-bin structure was structurally flawless — the concrete bins did not crack — yet it rotated as a rigid block while one edge dropped roughly nine metres into the ground and the opposite edge lifted clear of grade. The cause was not in the building at all. The shallow raft beneath it had loaded a deep, soft plastic-clay stratum past its ultimate bearing capacity, and the soil failed in general shear, flowing out sideways from under the foundation.

The elevator had been completed in the early autumn of 1913 for the Canadian Pacific Railway, built to relieve the Winnipeg yards at the peak of the grain harvest. It comprised five rows of thirteen cylindrical concrete bins, each about 28 metres tall and 4.4 metres in diameter, all carried on a single 600-millimetre reinforced-concrete raft measuring 23.5 by 59.5 metres, founded only about 3.6 metres below grade. The design rested on a plate-load test that, performed near founding level with a small plate, had measured only the firm upper clay and never reached the weaker stratum metres below that actually governed the foundation's strength.

Filling began bin by bin. The elevator behaved normally until it had taken on roughly 875,000 bushels — about 87.5 percent of capacity. Then settlement was noticed; within an hour it reached some 300 millimetres uniformly, and the whole structure began to lean west. The tilt grew through the night and the following day until the elevator stood near 27 degrees, one side embedded deep in the clay, the other reared into the air. The applied foundation pressure had reached roughly 290 kilopascals; the true bearing capacity of the two-layer clay profile, as later back-analysis showed, was below that.

The remediation became as celebrated as the failure. Rather than demolish a sound structure, the owners underpinned the 20,000-ton bin house, hand-dug wells beneath it, and jacked it slowly upright on new piers carried down to firm ground. Righted by October 1914 and standing about 3.6 metres lower than before, the elevator went back into service and is regarded as the textbook demonstration of ultimate bearing capacity — the case that taught the discipline that a foundation is only as strong as the deepest weak layer its pressure bulb can reach.

---

Timeline

1911

Construction begins

The Canadian Pacific Railway begins building a million-bushel reinforced-concrete grain elevator at North Transcona, near Winnipeg, to ease congestion in the Winnipeg yards during the grain rush.

1911–1913

Bin house erected

A bin house of 65 cylindrical concrete bins — five rows of thirteen — is raised, each bin about 28 m tall and 4.4 m in diameter, on a single 600-mm reinforced-concrete raft, 23.5 × 59.5 m, founded roughly 3.6 m below grade.

Pre-construction

Plate-load test only

The bearing value is set from a plate-load test near founding level reporting a high allowable pressure. No deep borings are taken; the soil is assumed homogeneous, matching the firm upper clay.

September 1913

Elevator completed

The structure is finished and ready to receive grain. The design foundation pressure at full load is on the order of 290 kPa, deemed safe against the tested value.

Early October 1913

Filling begins

Grain is loaded bin by bin. The elevator settles within expected tolerances and shows no distress through the early stages of filling.

18 October 1913

First movement

With about 875,000 bushels in store — roughly 87.5% of capacity — vertical settlement of the bin house is observed and grows to about 300 mm within an hour.

18 October 1913

Tilt begins

Settlement ceases to be uniform; the structure begins to rotate toward the west as the clay beneath one side yields and heaves out at the surface beside the foundation.

18–19 October 1913

Rigid-body rotation

Over roughly 24 hours the bin house tilts to an inclination near 27 degrees; the west edge sinks about nine metres while the east edge rises several metres above grade. The bins do not crack.

December 1913

Righting begins

The Foundation Company of Montreal and Vancouver is engaged. Workers hand-dig five-foot-diameter wells beneath the raft, install concrete piers carried to firm strata, and prepare to jack the structure upright.

Through 1914

Slow jacking

Using screw jacks with horizontal pushers, the structure is moved a few inches per day, excavating beneath the high side and lifting against the new piers.

14 October 1914

Elevator righted

About a year after the failure, the bin house is brought back to vertical and supported on the new piers, standing about 3.6 m (12 ft) lower than its original level. It is returned to service.

1951–1953

Modern analysis

Renewed soil investigation confirms a soft clay stratum beneath the firm upper layer; Peck and Bryant publish a back-analysis showing the failure was a classic bearing-capacity failure, cementing the case in geotechnical teaching.

The Build: A Tall, Heavy Structure on a Thin Raft and a Single Plate Test

The Transcona elevator was a creature of the wheat boom. The Canadian Pacific Railway needed storage to handle the harvest surge moving through Winnipeg, and it built big: a bin house of 65 reinforced-concrete cylinders arranged in five rows of thirteen, each bin standing roughly 28 metres tall and 4.4 metres across, with a combined storage of about a million bushels. The whole assembly was carried on one continuous reinforced-concrete raft, 600 millimetres thick and 23.5 by 59.5 metres in plan, set only about 3.6 metres below the surrounding ground.

That was a structure of unusual concentration. A grain elevator is, in load terms, close to the worst case a shallow foundation can face: a tall, rigid, very heavy block whose weight is dominated not by the structure but by the dense commodity poured into it, applied uniformly across a small footprint. At full load the raft pressed on the clay at something near 290 kilopascals. Whether the ground could carry that pressure was the single decisive question of the project — and it was answered with one small test.

The site rests on the lacustrine clays of glacial Lake Agassiz, the flat bed of an Ice Age lake that once covered much of Manitoba. Beneath a thin soft surface layer lies a firm brown clay, and beneath that a softer, weaker grey-blue plastic clay locally known as "blue gumbo." The fatal limitation of the plate test was geometric. A small plate stresses only the soil immediately beneath it; its zone of influence is shallow, confined to the firm upper clay. The 23.5-metre-wide raft, by contrast, pushed a pressure bulb deep into the ground — far enough to reach and overstress the soft stratum the plate had never felt. No deep borings were taken to find it. The designers assumed the clay was uniform, and built as if the strong layer continued downward forever.

The Failure Sequence: A Pressure Bulb Finds the Soft Layer and the Clay Flows Out

Filling proceeded uneventfully through the early autumn of 1913. The elevator settled, as all foundations do, but within tolerances that gave no warning. The ground was being asked, slowly, the question it would fail — but only the last increment of load would push it over the edge.

That edge arrived at about 875,000 bushels, roughly 87.5 percent of capacity, on 18 October 1913. Settlement was noticed and, within an hour, had grown to some 300 millimetres of nearly uniform sinking. Then the symmetry broke. The clay beneath the west side of the raft had reached its ultimate shear strength and could carry no more. In a general-shear bearing failure, the soil does not merely compress; it ruptures along curved slip surfaces that run down beneath the foundation, sideways, and back up to the surface beyond the edge, where the displaced clay heaves up in a bulge. Once that continuous failure surface forms, the foundation is no longer supported — it is riding on soil flowing out from under it.

That is exactly what happened. The west edge of the raft drove downward into the yielding clay while the soil beside it bulged upward, and the entire rigid bin house began to rotate westward about its long axis. The tilt grew through the night and into the next day, an irreversible rotation that no measure could arrest until the structure had found a new, deeply embedded equilibrium at an inclination of nearly 27 degrees. The west side had sunk some nine metres into the prairie; the east side stood several metres clear of grade. Throughout, the concrete bins remained essentially undamaged. Nothing in the structure had failed. The building was perfect and the ground had given way — the cleanest possible separation of a structural problem from a geotechnical one, performed at full scale.

The Reckoning: A Sound Structure Jacked Upright, and a Discipline Given Its Canon

The response was extraordinary for its time. The bin house, though grotesquely tilted, was intact and worth saving, and the owners chose to right rather than raze it. The Foundation Company of Montreal and Vancouver undertook the work from December 1913. Crews hand-dug wells about five feet in diameter beneath the raft and built concrete piers down to firm strata, creating fixed points to push against. Then, with banks of screw jacks and horizontal pushers, they coaxed the 20,000-ton structure back toward vertical at only a few inches a day, excavating under the high side as the low side was lifted. By 14 October 1914, about a year after the failure, the elevator stood plumb again on its new piers, resting roughly 3.6 metres lower than it had begun, and went back into service.

The diagnosis proved more consequential than the repair. At the time the theory of bearing capacity was immature; engineers had no general equation linking foundation width, depth and soil shear strength to the load that would cause a shear failure. Transcona supplied a full-scale, well-documented experiment in precisely that. Renewed investigation around 1951 confirmed the layered profile — firm clay over the soft "blue gumbo" — that the original plate test had missed, and in 1953 R. B. Peck and F. G. Bryant published a back-analysis showing the applied pressure had simply exceeded the ultimate bearing capacity of the weaker deep layer. The numbers closed: the load that destroyed the foundation matched the load that bearing-capacity theory, by then maturing under Karl Terzaghi's framework, predicted it should. Transcona became the empirical anchor for the concept of ultimate bearing capacity in soft clay. The error it named was specific and timeless — sizing a wide foundation from a small, shallow test that never reached the layer the foundation's own pressure bulb would govern.

---

Contributing Factors

01

Bearing capacity judged from a shallow plate-load test

The allowable pressure was set from a small plate loaded near founding level. A plate stresses only the soil directly beneath it, to a shallow depth; the firm upper clay it measured was not the layer that would govern a 23.5-metre-wide raft. The test answered the wrong question — local strength at the surface, not the deep strength that mattered.

02

An undetected soft stratum within the pressure bulb

Beneath the firm brown clay lay a weaker grey-blue plastic clay, the "blue gumbo" of Lake Agassiz. A narrow plate never reached it, but the wide raft pushed a pressure bulb deep enough to overstress it. The strength of the foundation was set by the weakest layer inside that bulb, and that layer was never found or tested before construction.

03

No deep borings; the soil assumed homogeneous

No subsurface investigation was carried out to characterise the clay with depth. The design treated the ground as uniform, extrapolating the strong upper layer downward indefinitely. The most heavily loaded element on the site — the deep clay — was the one element no one had examined.

04

A concentrated, predictable, near-uniform load on a shallow raft

A grain elevator applies an enormous, evenly distributed pressure over a small footprint, dominated by the stored commodity rather than the structure. Carried on a thin raft only 3.6 metres deep, that load offered the soft layer no relief from foundation depth or width and drove it straight to a general-shear failure once filling neared full.

05

General-shear failure in soft clay, expressed as rigid-body tilt

Because the bin house was extremely stiff and the clay failed along a continuous slip surface on one side, the structure did not deform or crack — it rotated as a single rigid block. The soil heaved out beside the foundation as the raft drove in, producing the near-27-degree tilt that left the building intact and the ground destroyed. ---

Aftermath

No one died, and the structure itself was never the problem — which is precisely why Transcona endured as a teaching case rather than a tragedy. The bin house was underpinned, jacked upright over the course of 1913–1914, and returned to service about 3.6 metres lower than it began, a feat of remediation still cited in its own right. The deeper legacy was theoretical. Transcona became the full-scale field test that the emerging science of soil mechanics needed: a heavy, well-instrumented foundation that failed cleanly in general shear, with a known load and a back-analysed soil profile. It supplied empirical confirmation for the ultimate-bearing-capacity theory that Karl Terzaghi and successors were formalising, and the Peck–Bryant analysis made it the standard worked example in soil-mechanics texts worldwide. It hardened a now-universal requirement of geotechnical practice: that foundation design rest on borings deep enough to sample every stratum within the pressure bulb, and that bearing capacity be governed by the weakest layer the structure can reach — never by a shallow plate test. Transcona is the byword for a single proposition: a building can be flawless and still fall over, because the structure is only ever as strong as the ground no one investigated.

---

Lessons

Investigate the soil to the full depth of the pressure bulb, not the founding level: a wide foundation stresses soil far deeper than a narrow one, so borings must reach and test every stratum the load can influence.
Never size a large foundation from a small plate test; scale changes the depth of influence, and a plate can report safe ground while the real foundation overstresses a weak layer the plate never touched.
Let the weakest layer in the profile govern bearing capacity, not the layer you happen to found on; assume the ground is layered until borings prove otherwise, and design for the softest clay within reach of the load.
Remember that a sound structure can still fail through the ground beneath it: distinguish a geotechnical failure from a structural one, because the fix lies in the soil, not the steel or concrete.
Watch settlement as the first warning, and act before it turns to tilt — uniform sinking that suddenly becomes one-sided rotation signals a shear failure already underway, and once the slip surface forms the foundation cannot be saved in place. ---