Building architecture for refugees is often a job done under intense constraints and rapidly changing conditions. It’s not an easy task, but right now, more than 117 million people depend on those builds.

That figure comes from UNHCR’s latest Global Trends report: by the end of 2025, roughly one in every 70 people on Earth had been forced from their home.

Refugee shelter design is where that need meets the drawing board. It is even harder than it looks much of the time (and it already looks hard).

Architecture for refugees tests speed, dignity, culture, and climate all at once.

Two Building Green guests have worked inside this world: Fabiano Sartori, an architect and planner with the Danish Refugee Council (Ep. 14), and Jess, a disaster manager who has deployed to crises across three continents (Ep. 10).

Their experiences and insights could completely change how you think about architecture for refugees.

In this post, we’re following their lead, from minimum standards to the growing shift toward reusing the buildings cities already have.

 

What Refugee Shelter Design Must Solve

Start with the constraints. After all, these can be strict.

Sartori described responses where hundreds of thousands of people arrive in months. Bangladesh received roughly 700,000 people in one such influx, for example, and there was no time for planning as architects know it.

The humanitarian sector answers these constraints with minimum standards. The Sphere Handbook, the closest thing to a global rulebook, sets the floor for refugee shelter design:

• 3.5 square meters of covered living space per person in warm climates, more in cold ones
• A minimum ceiling height of two meters, with ventilation
• Site planning that accounts for cooking, sanitation, drainage, and firebreaks
• Space standards that flex to reflect cultural and social norms

And yes, that first number is correct. 

Three and a half square meters is a parking space, roughly, and reaching even that for a hundred thousand people in a short space of time is a wild achievement.

The biggest problem is time.

Sartori calls camps “permanent temporary,” assembled in days as emergency responses, then inhabited for decades.

The World Bank confirms that two-thirds of refugees have been displaced for more than five years, overwhelmingly hosted in developing countries.

And an earlier World Bank–UNHCR study, published via ReliefWeb, put numbers on where that burden falls: developing countries host 89% of the world’s refugees and 99% of its internally displaced people.

The places with the fewest resources carry the heaviest load, often in climates growing harsher by the year.

So the central paradox of architecture for refugees is designing something fast enough for an emergency and good enough for a childhood.

 

Two Famous Answers and What They Teach

 

A handful of projects have defined public imagination on this topic, and offer lessons.

The first is Shigeru Ban’s paper architecture.

After the 1994 Rwandan crisis, Ban proposed shelters built from cardboard tubes to the UN’s refugee agency, which hired him as a consultant. And his Paper Log Houses, with beer-crate foundations and paper-tube walls, went on to shelter displaced families from Kobe to India.

The Pritzker jury cited that humanitarian work when awarding him architecture’s highest honor in 2014.

The second is Better Shelter, the flat-pack unit born from a partnership between the IKEA Foundation and UNHCR.

By Better Shelter’s own count, some 90,000 of its units have housed people across 80 countries, with a goal of reaching one million people by 2030. The design logic:

• A lockable door: safety, especially for women and children, that a tent cannot offer
• A solar panel and lamp, since camp electricity is scarce
• Assembly by four people in about four hours, no specialist tools required
• A lifespan of years rather than the months a tent survives

Yet, even this celebrated product needed a redesign after fire-safety concerns paused deployments in 2017, proof that refugee shelter design is never a finished object, only an iteration.

There’s a parallel here to what we’ve written about renewable materials: the material is never the whole answer. Often, the system around it determines the extent to which it works.

 

Culture is a Structural Material

After hurricanes flattened parts of the US Virgin Islands, Jess flew in expecting total destruction. But she found one building type still standing.

Circular homes, built in the tradition of the African rondavel by families who had lived there for generations, survived because the wind had no corners to grip.

The resorts around them were gone.

Indigenous knowledge, she argues, is load-bearing: people build the way they do because they understand where they live.

She has also seen the opposite: NGOs delivering prefab homes in East Asia that families simply refused to occupy because the layouts ignored how those communities actually live, with large shared spaces and separate sleeping quarters.

An empty shelter is a failed shelter, whatever its engineering says.

Her favorite counterexample comes from Nepal: architects working alongside communities to design bamboo homes, jointed without metal hardware, earthquake-resilient, shaped around how residents wanted to live – and, she adds, beautiful.

This makes me think of a saying a builder once shared with me: people don’t live in relation to structures but through their habits.

Good architecture for refugees requires acknowledging and working with these habits.

The same participation principle shows up in shelter case studies collected by the humanitarian shelter community, and it echoes what we found exploring biophilic design: buildings do their best when they fit real human needs.

 

Conceiving of the City as a Shelter

 

Now, the twist.

The iconic image of displacement is a camp.

But Sartori points out that more displaced people live in cities than in camps, negotiating housing, work, and services alongside host communities.

Research on applying Sphere standards in urban settings found that displaced families often refuse relocation to formal sites, because moving away from urban districts means losing income and schools.

So the frontier of refugee shelter design is increasingly not a product at all.

It’s adaptive reuse.

Sartori advocates converting existing buildings into shelter wherever possible, for reasons that stack up fast: construction impacts avoided, speed gained, and, perhaps most important, displaced families woven into neighborhoods with access to jobs and services instead of isolated at the edge of town.

Every reused structure also sidesteps the upfront emissions of new construction, an idea we’ve explored before in our post on embodied carbon vs operational carbon.

Sartori also offered a forward-looking example from Brazil’s response to the Venezuelan crisis:

Emergency shelters that could later serve as transitional housing for homeless Brazilians, one investment serving two vulnerable groups in sequence.

That’s planning a building’s second life before its first one ends, thinking that would improve plenty of conventional projects too, as our smart cities guide argues.

The barrier, Sartori is careful to note, isn’t a technical one.

Governments often resist making shelter too good or too integrated, fearing permanence. This is firmly a political constraint, and it’s one no amount of clever refugee shelter design can engineer away, a reminder of how much building rules and codes shape outcomes everywhere.

Principles to Carry Home

Distilled from both conversations, the fundamentals of good architecture for refugees look something like this:

• Design with, never for: Participation is what separates occupied shelters from abandoned ones
• Plan the second life: Assume decades, and design the upgrade path, much as a passive house retrofit upgrades a home in deliberate stages
• Reuse before you build: Existing buildings mean faster shelter, lower impact, and real integration
• Respect climate and culture equally: Insulation physics matter, the same physics behind EnerPHit certification, and so does where a family expects to cook
• Anticipate, rather than respond: Sartori’s sector is shifting toward acting on climate data before disasters strike, since climate pressures now intersect with displacement across the regions his organization serves

None of this stays confined to humanitarian work either, accounting for housing that goes up fast, sips energy, survives storms, and honors how people live.

All of this fits the brief for the whole industry this century, from net-zero homes to city districts.

The refugee context just forces us to face the facts of where we’re headed, removing the luxury of pretending that building for these conditions is optional.

At Building Green, we’ll keep having these conversations because the hardest constraints produce the clearest thinking.

Shelter is the most bare-bones work of architecture. And for 117 million people, it’s also the most urgent.