The first time he drove through Port-au-Prince, Haiti, back in 2004, Scott Hamel had a singular reaction: “Wow! An earthquake would devastate this place.”

Hamel was there as part of an Engineers Without Borders project. He immediately focused on the two ways locals built their roofs.

The least popular are made of tin; they blow off, leak and are deafening in rainstorms. The preferred roofs are thick concrete slabs, precariously perched atop rather crumbly masonry block walls.

These thick roofs are especially popular because builders often use them as the floor for a second story. Five or six inches thick and loaded with reinforcing steel, they are heavy and expensive, and require massive amounts of human labor.

Typically, workers mix the concrete on the ground and manually lift it up in five-gallon buckets. “Can you imagine doing that all day?” Hamel asks. Not only that, but the cost of these roofs — lots and lots of cement — is often borne by well-meaning U.S. churches and charities that offer financial support for better housing.

All those minuses were percolating in Hamel’s engineering brain as he pondered a better roof. But the best argument for a safer design came at about five o’clock on the evening of Jan. 12, 2010. A catastrophic magnitude 7.0 earthquake with an epicenter just 16 miles west of Port-au-Prince struck Haiti, collapsing structures, injuring 300,000 and killing an estimated 230,000. Those heavy concrete roofs were a massive part of the devastation.

Fast forward to December 2013. By then, Hamel had joined the civil engineering faculty at UAA and moved to Anchorage, where he and his wife are busy raising their small children. Hamel’s last trip to Haiti was in 2010.

But he hadn’t forgotten the Haitian roof problem, one repeated around the globe, from Africa to South and Central America — any place where housing is insufficient. He applied for seed money from UAA’s Innovate fund to try his hand at a new idea: the concrete, barrel-shaped shell as a potential new roof system in Third World countries.

His new idea is really an old one employed in the United States up to the middle of the last century, before high labor costs and cheaper material costs shifted this design to the sidelines. But architects are still drawn to the rounded simplicity of those rounded concrete shells, demonstrated in the spectacular roof of the Sydney Opera House.

In 2013, Hamel got $10,000 in seed money to design and evaluate a “thin-shell, latex-modified concrete barrel roof unit.” The latex is leftover paint that is difficult to recycle. It’s also a sophisticated polymer that, when added to concrete, increases its flexibility.

Hamel used the money to buy supplies, design and build a fiberglass mold for the shell, cast the shell and then build the tools to test its structural integrity.

He hired undergraduate civil engineering student Nathaniel Cox to be a part of his design team. Besides the single physical shell, they also designed the reusable mold, tested the concrete shell’s integrity and developed computer models that allow them to calculate shear, compression and tensile strength on units with different dimensions.

Their single physical unit was six feet long, three feet wide, one foot high and ¾ of an inch thick with a mesh layer in the middle. Just looking at it, Hamel already thinks eight feet might be a better length. To make a roof, a series of these barrel units would be joined together over a building.

Why would these work in Haiti? Because they are cheaper to make than concrete slab roofs; many can be made from the reusable mold; they can be made on the ground when residents have time, like when their crops are growing. And, because they are so much lighter, completed and cured shells can be easily carried up a ladder for installation. In an earthquake, crumbling concrete shells may still cause injuries, but they are much lighter and less likely to cause death.

Hamel has made some conservative material cost estimates and figures the thin shells can be made for $2-$3 per square foot, compared to $6-$10 per square foot for a comparably sized five-inch thick concrete slab. The savings comes in less cement and steel needed for the finished product.

I asked Hamel what comes next. He’s searching for funding to run a pilot project in a location like Haiti. He thinks the U.S. Agency for International Development (USAID) might be a prospect, or maybe some of the charities that previously funded the ill-fated thick-slab-roofed homes that crumbled in the 2010 earthquake.

I asked about applications in rural Alaska. More cold-climate and snow-load testing would be required, Hamel said, but the barrel shell roofs are very economical to make in remote locations. They could be made on-site and cured in winter in above-freezing warehouses or cellars, ready for the summer construction season.

Kathleen McCoy works at UAA, where she highlights campus life through social and online media.