[Content note: this talk includes discussion of loss of life.]

On 6th November 1917, the SS Mont Blanc sailed into Halifax containing a startling cargo of explosives and munitions, destined for Allied forces in World War I. On the same morning, the SS Imo sailed out, empty and in a hurry.

In high traffic, constricted areas like this, marine traffic has rules to follow much like vehicular traffic in a city. In particular: drive on the right, and stick to the speed limit. The Imo, going faster than recommended, found itself (in effect) driving on the left at the
narrowest point in the channel, just as the Mont Blanc approached. The Imo crew was unaware of the cargo in the Mont Blanc.

A funny thing about marine collisions is that they happen remarkably slowly. Each ship took evasive manoeuvres, but the momentum of each and the constricted area meant that the ships still collided, the Imo - engines full reverse - sideswiping into the Mont Blanc. This knocked over some of the benzol barrels in the Mont Blanc's cargo, but did not ignite them. When the Imo withdrew, though, the sparks generated by the metal hulls grinding against each other lit the fire. Crowds gathered on the shores to watch the
spectacle of the Mont Blanc on fire.

The Mont Blanc was the largest man-made explosion to date.

The blast killed 1600 people instantly, and nine thousand were injured. A piece of the ship's anchor was found over two miles away.

This is... heavy stuff, especially for a networking conference. What I plan to focus on is not the explosion itself but what happened before, and what happened after.

When we look at the run-up to the explosion, we can learn a lot by teasing apart the contributing factors to the explosion, finding what mistakes were made, and understanding why they were made. This was an exceptional situation, not least because it was taking place in wartime, when normal safety rules might be suspended. The Imo was proceeding with haste, but lacked the information that might have caused it to slow down; those who had the information did not or could not communicate it. This was a classic example of Dr. J Reason's "Swiss Cheese" model of accidents.

In this talk, we'll learn how to avoid and deal with accidents by looking at how other industries have responded to theirs. We'll learn the limitations of "root cause" as a concept, understand more about why mistakes get made, and how to spot the signs of scapegoating and manage them.