The Sticky Physics of Disaster: How 2.3 million gallons (8.7 million liters) of molasses killed 21 people in 1919, exposing the dangerous duality of non-Newtonian fluids.
When a 50-foot steel tank ruptured in Boston’s North End on January 15, 1919, it unleashed not just molasses, but a fundamental lesson in materials science. The conventional wisdom holds that industrial accidents stem from obvious mechanical failures—ruptured pipes, collapsed structures, or runaway machinery. Yet the data from this peculiar disaster reveal a more counterintuitive reality: the most dangerous engineering failures often emerge from the unexpected behavior of seemingly benign materials under stress.
The tank held 2.3 million gallons (8.7 million liters) of crude molasses, a substance so viscous that workers reportedly used it as a makeshift ladder when maintenance was needed. But when the tank failed, the molasses behaved like a compressed spring, creating a wave that moved at 35 miles per hour (56 km/h) and reached heights of 25 feet (7.6 meters). The result was 21 deaths and 150 injuries, with the sticky flood destroying buildings and derailing a train. The evidence suggests that this disaster was not merely a structural failure, but a profound misunderstanding of material properties.
The initial engineering consensus viewed molasses as a stable, predictable substance—thick enough to store in massive quantities, viscous enough to resist flow. Early 20th-century industrial data showed molasses tanks operating safely across North America, with the Purity Distilling Company in Boston maintaining its facility since 1915. The material’s apparent stability made it an ideal industrial storage medium, cheaper and safer than alternatives like petroleum.
However, a deeper analysis of fluid mechanics data reveals a more complex picture. Molasses is a non-Newtonian fluid, meaning its viscosity changes dramatically under stress. While it behaves like a thick syrup under normal conditions, the immense pressure inside the tank—estimated at 250 pounds per square inch (1.7 MPa)—transformed it into a far less viscous substance. The data indicate that this pressure drop reduced the molasses’s effective viscosity by up to 90%, allowing it to flow with the speed and force of water.
The tank itself compounded this material failure. Engineering reports show the steel plates were only 3/16 inch (4.76 mm) thick, far below the 1-inch (25.4 mm) minimum recommended for such pressures. The 1.5 million rivets holding the tank together were spaced too far apart, creating weak points that failed under the combined stress of internal pressure and thermal expansion. The evidence suggests that temperature fluctuations during the unusually warm January day may have contributed to the final rupture, with metal expansion creating additional stress on already inadequate fasteners.
This pattern is not universal. Some industrial accidents do result from truly unforeseeable events, such as the 1979 Three Mile Island incident where multiple system failures cascaded unpredictably. Yet for the vast majority of documented industrial disasters, the data indicate that root causes involve known physical principles that were inadequately addressed. The Boston case reveals how human assumptions about material behavior can amplify mechanical shortcomings into catastrophe.
The implications of this analysis extend far beyond historical curiosity. As industries increasingly work with complex materials—from nanomaterials to bioengineered substances—the data suggest that material testing under real-world conditions becomes ever more critical. The challenge for modern engineers is to treat material properties not as constants, but as variables that can transform dramatically under stress. The historical record indicates that when material science is oversimplified, the results can be as sticky as they are deadly.
The ruptured tank in Boston’s North End may be gone, but its data continue to speak. In an age of advanced materials and computational modeling, the evidence suggests that the most valuable engineering tool remains rigorous empirical testing. The next industrial disaster, the data indicate, will likely be prevented not by more sophisticated theories, but by remembering that even the most familiar substances can harbor unexpected dangers.
The Boston Molasses Flood illustrates how material properties can behave unexpectedly under stress, and how structural inadequacies can amplify these effects. The engineers knew molasses was viscous but didn’t understand that it becomes far less viscous under sudden pressure—classic non-Newtonian behavior.
Unlike visible structural or material failures, our next disaster was born from a failure of information itself—a single, invisible error on a surveyor’s map.
In our next post, we’ll examine the Lake Peigneur Disaster, where a 400-foot triangulation error turned a peaceful Louisiana lake into a massive whirlpool that swallowed barges, docks, and an entire island in just three hours. Continue to The Lake Peigneur Disaster →
External Sources
- Petroski, Henry. To Engineer Is Human: The Role of Failure in Successful Design. St. Martin’s Press, 1985.
- Perrow, C. Normal Accidents: Living with High Risk Technologies - Updated Edition. (Princeton University Press, 2000). doi:10.1515/9781400828494.
- “The Great Molasses Flood.” History.com