Building for the Next Hurricane, Not the Last One

Every season I have some version of this conversation: the last house came through fine, so why does this design need straps, rated openings and an engineered roof detail?

It is a fair question and it deserves a real answer rather than reassurance. The answer is that survival is not evidence of adequacy. A house that stood through a tropical storm has been tested against that storm's wind field, from that direction, with that rainfall, at that time of day. It tells you very little about the next one.

Barbados makes this trap especially easy to fall into, because the island sits at the eastern edge of the belt and has been spared direct major-hurricane landfall for a long time. That record produces a dangerous local intuition: that hurricanes are something that happen to other islands.

Two events are worth holding in mind, because neither was a worst case and both did real damage.

In October and November 2010, Tropical Storm Tomas brought down trees and utilities across the island, damaged roofs, and caused enough structural damage along the Bridgetown waterfront that engineering assessment and repair work kept a stretch of Wharf Road closed for weeks.

In July 2021, Hurricane Elsa passed close to Barbados as a compact system near the lower end of the hurricane scale. According to reporting from Barbadian authorities and regional agencies, roofs on well over a thousand homes were damaged, with a number of houses destroyed outright — and much of that damage occurred in a matter of hours during a morning passage.

Before → Now → Proposed

BEFORE

NOW

Barbados' recent building damage has come overwhelmingly from roofs, cladding, openings and coastal edges — not from walls falling down.

I want to state this precisely rather than dramatically. The IPCC's Sixth Assessment Report concludes that the proportion of tropical cyclones reaching the most intense categories has likely increased globally over recent decades, and that heavy rainfall associated with tropical cyclones is expected to increase with further warming. What it does not conclude is that there will simply be more storms everywhere.

For a designer in Barbados, the practical reading is precise: the upper end of the intensity range is more available than it used to be, and more water may arrive with it.

Wind lifts roofs; it does not press them down. Suction on the upper surface combines with internal pressure the moment an opening fails, and the roof is pulled and pushed upward at the same time. Everything else follows from that.

The most common weak point I find in existing Barbadian houses is not the roof structure itself. It is the last metre of it: a well-framed roof sitting on a wall with no engineered tie, or with straps installed after the fact into concrete that had already cured around nothing.

Wind damage makes the photographs; water does most of the financial damage. Rain driven horizontally finds every unsealed junction, and after roof damage a house can take on more water in an hour than in years of ordinary weather.

In practice that means designing the wet aftermath as carefully as the wind event: oversized gutters and downpipes, drainage that moves water away from the building rather than into a low courtyard, electrical outlets and distribution boards positioned above likely water levels, and floor and wall finishes at ground level chosen so they can be dried rather than replaced.

I would not romanticise the old buildings. Timber, hand-driven fixings and loose block foundations have hard limits, and storms took many of these houses. But the geometry is still instructive, and it is free: a simple, steep, compact roof with modest overhangs asks far less of its fixings than a complicated one with wide unsupported eaves.

Resilience in Barbados is not exotic engineering. It is a small number of details, done properly, in the right order: a continuous load path, protected openings, a simple roof, and drainage that assumes far more water than the average year delivers.

What I take from the last two decades of regional storms is that the difference between a bad week and a lost year is usually decided at design stage, for a fraction of the cost of the repair. That is why I would rather spend a client's money on connections and openings than on square footage.