Nearly 100 years ago, a Boston neighbourhood was swamped by a tidal wave of molasses. BBC Future looks at a design defect that created a bizarre – and deadly – flood.

As we go about our daily lives, we have a fairly reasonable expectation that the objects around us will continue to maintain their structural integrity. That subway tunnel will keep its shape; that bridge will stay up; that elevator to the 39th storey isn't going to plummet earthwards, unless you ask it to.

Engineers tasked with ensuring public safety do a remarkably good and thorough job. But there have been moments when all was not well, and the results have been shocking – for instance, when what appear to have been design failures led to one of the oddest man-made disasters of the last century: The Great Molasses Flood of 1919.

An enormous steel tank containing 2.3 million gallons (8.7 million litres) of molasses was a familiar landmark in Boston’s North End, looming above a playground, a fire station, and private homes. The molasses came from the West Indies to be converted into industrial ethanol by the tank's owner, the US Industrial Alcohol Company, who had been selling it for use in munitions for the Great War.

The tank had been built in a bit of a hurry. As Stephen Puleo, author of the book Dark Tide, says, it leaked dripping sticky black goo from its riveted seams. Local children flocked to sneak a treat from the drippage, but it made at least one of the company's employees nervous.

The tank’s presence had been uneventful for the four years since it had been built. But on 15 January 1919, all hell broke loose.

North End was going about their business as usual. Then, at 12:40pm, there was a sound like machine gun fire and a terrible wrenching groan. “The explosion came without the slightest warning,” the Boston Globe reported the next day. “Once the low, rumbling sound was heard no one had a chance to escape.” A wall of molasses reported to be 25-feet-high (7.5 metres) came heaving away from the collapsing tank.

The wave flattened buildings. It picked people up and slammed into the elevated rail tracks nearby with such force that the metal crumpled. “The buildings seemed to cringe up as though they were made of pasteboard,” the Globe wrote. The curving sheets of steel that had made up the tank were borne along on the tsunami of viscous liquid sugar, crashing through everything that stood in their way.

Waist-deep molasses settled and thickened over the neighborhood, and animals and people struggled to free themselves

“Anthony di Stasio, walking homeward with his sisters from the Michelangelo School, was picked up by the wave and carried, tumbling on its crest, almost as though he were surfing,” wrote Edwards Park in a 1983 Smithsonian feature on the disaster. “He heard his mother call his name and couldn’t answer, his throat was so clogged with the smothering goo. He passed out, then opened his eyes to find three of his sisters staring at him... They had found little Anthony stretched under a sheet on the ‘dead’ side of a body-littered floor.”

The aftermath was unspeakable. Waist-deep molasses settled and thickened over the neighborhood, and animals and people struggled to free themselves. When the bodies were recovered, the death count was 21, including Maria di Stasio, Anthony’s 10-year-old sister, and a 65-year-old housewife who had died as her house collapsed around her. More than 150 others were injured.

There was much speculation at the time that the tank might have been blown up by Italian anarchists. Ultimately, however, the US Industrial Alcohol Company was found liable for the disaster, after six years of litigation. At the time, one of many experts who testified in the case noted that the walls of the tank were thinner than the plans called for and used fewer rivets than would have been standard.

Over the last few years, Ronald Mayville, a Boston-based structural engineer, has been investigating what exactly went wrong in more detail, using modern techniques. He makes his living doing the kinds of tests and calculations required to make sure such structures built today are safe. And by poring over 20,000 pages of testimony from the trial, Mayville has found a few telling details.

One of them concerns the tank’s rivets, which probably made that machine-gun-like sound when they popped free. When you punch holes in steel to thread rivets through, it makes the surface of that metal around the hole brittle. To compensate, you usually shave off that layer, so only the metal surrounding it is full-strength. But it is possible that in their haste, the tank’s builders did not take this step. Add to that, any time you cut a hole in a sheet of metal under tension, stress builds up around the hole.

There was no need for a bomb to go off to cause sudden structural failure – the potential for it was built into the design

The tank was also pierced with a hatch called a manway. In addition to the stress concentrated around the hole of the manway opening, the outer part of the hatch was riveted to the steel in such a way it created a great deal of stress on the top-most rivet hole. And when the remains of the tank were inspected, Mayville relates, there were herringbone patterns leading across the metal to the top and bottom of the manway.

“We know that for the fracture of brittle steels, the herringbones make little arrows that point back to the location where the error started. There was a major fracture that went up to the manway, one underneath,” he says. That was probably where the tank tore, and contrary to what US Industrial Alcohol’s lawyers maintained, there was no need for a bomb to go off to cause sudden structural failure. The potential for it was built into the design.

The steel plates were also much thinner than they should have been. But the tank pieces, oddly enough, were not made by some fly-by-night operation. They were constructed by a reputable firm, Hammond Ironworks, that actually had experience designing tanks that held molasses.

Mayville says that it’s possible a fatal miscommunication might have occurred early on. “My suspicion,” he says, “is that the designers back in the shop thought the tank was going to hold water.” Molasses is far denser than water, and thus exerts more pressure on tank walls. If the designers thought the tank was going to be holding water, it could explain a lot.

Since the Molasses Flood, building codes and standards have evolved. Boilers made the same way as the tank suffered so many explosions that people changed the way they were built. Every disaster still provokes the question of what exactly was at fault, and the standards for big tanks today reflect all the accumulated knowledge of those investigations.

“In society it’s perfectly reasonable for you to assume that things have been designed correctly,” Mayville says. The problems mainly come about when a structure’s used in a way the builders didn’t intend, or when someone cuts corners, or when there’s been a miscommunication – or, perhaps as may be the case in the tragedy of The Great Molasses Flood, all at once.

Join 600,000+ Future fans by liking us on Facebook, or follow us on Twitter, Google+, LinkedIn and Instagram.

If you liked this story, sign up for the weekly bbc.com features newsletter, called “If You Only Read 6 Things This Week”. A handpicked selection of stories from BBC Future, Earth, Culture, Capital, Travel and Autos, delivered to your inbox every Friday.