Thermal mass passive heating systems use dense materials like dirt, stone, and concrete to absorb and store heat from a fire, then slowly release it over many hours, providing sustained warmth without continuous fuel consumption. The system works because heavy, dense materials have high thermal mass and can hold heat much longer than air, which loses heat quickly. This principle, used in ancient Korean ondol systems for over 1,000 years, allows a single 2-hour fire to maintain comfortable temperatures (around 65°F) for 8-12 hours after the fire has completely extinguished.
Deep Dive
Prerequisite Knowledge
- No data available.
Where to go next
- No data available.
Deep Dive
Homeless Peers Called Her "Stupid" Heating Trick Dangerous - Until She Hit 65°F With No Fuel at AllAdded:
65°.
That was the number on the meat thermometer she'd stolen from a Goodwill bin outside the San Mateo store on Central Avenue. The one with the cracked dial and the needle that stuck unless you tapped it. She tapped it 65° F and the fire had been out for 9 hours. Not banked, not smoldering out. Cold ash in a pit 12 ft from where she sat. Her name was Gail Pressman. She was 46 years old.
And on the morning of February 11th, 2024, she was the warmest person sleeping outside in Albuquerque, New Mexico. And not a single soul around her understood why, including, if she was honest, herself. Not completely. She knew what she'd built. She knew it worked. But the why of it, the full why, the physics underneath the dirt underneath the floor, underneath her body, some of that she was still putting together even then, even sitting there at 65° with her boots off and her socks drying on a cinder block ledge she'd mortared level for exactly that purpose.
Before any of this makes sense, before the trench and the channel and the chimney pipe and the argument she had with a man named Dale about whether she was building her own gas chamber, you need to understand where she was.
Not just the city, the actual spot.
South of Gibson Boulevard, east of the railard, there's a strip of land between a concrete drainage ditch in the backside of an abandoned roofing supply warehouse. The warehouse had been closed since 2019. The owner died. The family fought over the estate. Nobody boarded the windows for 3 years. And by the time Gail showed up in late September 2023, the lot behind it had already collected a small camp. Seven people, two tents, one van with no engine, a tarp leaned two, and a woman named Charlene who slept in a broken hot tub shell she dragged out of the warehouse loading dock. The ground was packed dirt and gravel, patches of goathead weed, a chainlink fence along the drainage ditch with three different holes cut in it.
The warehouse wall on the north side blocked wind from that direction, which was the only structural favor the landscape offered. Gail had been homeless for 14 months when she arrived there. Before that, she'd worked receiving at an Amazon fulfillment center in Rio Rancho until she tore her rotator cuff, pulling a jammed conveyor belt free with her hand instead of waiting for maintenance. Workers comp paid for 4 months. The surgery went fine. The recovery took longer than the payments lasted. She lost the apartment on Wyoming Boulevard in January 2023 and spent the spring couch surfing with her sister in Bellon until the sister's boyfriend said Gail had to go. She didn't drink much. She didn't use. She had a chipped front tooth from biting into a frozen Snickers bar in 2021.
Freckles that had faded to almost nothing after years of sun and wind. and strawberry blonde hair that had grown past her shoulders because she couldn't afford to cut it and refused to let anyone at the shelter do it with kitchen scissors again. She parted it crooked, always on the left side, and never fixed it. A thin scar ran across her left palm from a box cutter that slipped at the warehouse. She was 5'4", and she'd lost about 15 lb she couldn't afford to lose.
She had no special training, no engineering background, no survivalist phase in her 20s. What she had was a phone with a cracked screen and free Wi-Fi at the McDonald's on Central and a question that wouldn't stop bothering her. The question was this, why does a campfire only warm the side of you that faces it? This sounds like something a child asks. It is something a child asks, but Gail asked it as an adult with a specific adult problem. which was that campfire heat goes in every direction at once and the vast majority of it goes straight up into the sky and the part of it that reaches your body only reaches the surfaces directly exposed to the flame which means your front is hot and your back is freezing. She'd been sitting around Dale's fire pit in early October, maybe the fifth or 6th of the month, turning herself like a rotisserie chicken every few minutes, thinking about how stupid this was. Not stupid in an abstract way, stupid in a math way.
The fire was burning actual wood that people had spent actual effort dragging to camp. And the amount of that heat reaching her body was, and she was guessing here, maybe 5%. Maybe the rest was going straight into the Albuquerque sky where nobody needed it. She said this out loud. Dale, who was 58 and had been on the street for 6 years and who had opinions about everything, told her that was just how fire worked. You sit near it, you get warm, you go to bed, you get cold. That's life. And that was true. That was exactly how fire had worked for Gail for 14 months and for Dale for 6 years and for every other person in that camp for as long as they'd been outside. Fire is a thing that happens in open air. And open air takes the heat before you can. But that night, Gail couldn't sleep. And when she couldn't sleep, she looked at her phone and she typed something into Google that she felt sort of embarrassed about even as she was typing it. She typed how to keep heat from a fire without being next to the fire. The first results were camping tips, hand warmers, heated blankets, useless stuff, battery powered, store-bought, none of it applicable to her life. She scrolled past all of it. Then she hit a page about something called a rocket mass heater. She didn't know what that was.
The page had a drawing of a fire going into a tube going into a bench made of mud, and the bench stayed warm for hours after the fire went out. She read the page twice. She didn't understand half of it, but one sentence stuck with her, and she remembered it weeks later, almost word for word. The sentence said something like, "The heat is stored in the mass of the bench, and the mass releases it slowly." The way a parking lot stays warm after sunset because the asphalt absorbed the sun all day. That she understood. She'd stood on hot parking lots. Everyone in Albuquerque has stood on a hot parking lot at 7:00 p.m. in July and felt the heat coming up through your shoes from asphalt that had been baking since morning. The sun is gone. The heat is still there. The parking lot is a battery and the sun charged it and now it's discharging through the soles of your feet. She kept reading down below the rocket mass heater stuff. A link took her to a Wikipedia article about something called on doll. She tapped it. The article was about Korean houses, old Korean houses, hundreds of years old, where they built the kitchen lower than the bedroom and ran the smoke from the kitchen fire through stone channels under the bedroom floor. And the floor stayed warm all night from heat stored in the stone, and the smoke went out a chimney on the other side of the house, and nobody inside breathed any of it. Gail sat in her tent reading about Korean floor heating at 2 in the morning on a Tuesday in October and something clicked that she couldn't uncclick. The fire doesn't have to be where you are. The heat doesn't have to come from the air. The ground can hold it. The ground wants to hold it. The ground is better at holding heat than the air is because the ground is heavy and dense and air is neither of those things. And if you can get the heat into the ground, the ground will give it back to you all night long. the same way a parking lot gives back the sun. She didn't sleep that night. She read until the McDonald's Wi-Fi dropped her at 4:00 a.m. And then she lay in her tent and stared at the ceiling and thought about trenches. The shape of it was simple, a fire in a pit outside away from the shelter. A trench connecting the pit to the space under her floor.
Cover the trench with flat stones or concrete. run it under the shelter and pipe it out the other side through a chimney tall enough that the rising hot air would pull new hot smoke through the underground channel. The smoke heats the dirt on its way through. The dirt holds the heat because dirt is heavy and heavy things hold heat. The fire burns outside. The smoke never enters the room. And when the fire dies, the ground keeps radiating. Cinder blocks, which she'd already noticed piled behind the warehouse, do the same thing. She told Dale about it the next morning. Dale said she was going to kill herself with carbon monoxide. That was a reasonable thing to say. People die from smoke in enclosed spaces all the time. Dale knew about a guy named Marco who died in a tent two winters back. Brought a charcoal grill inside for warmth and never woke up. Everybody in camp knew about Marco. Dale told the story again, standing by the morning fire, looking at Gail like she'd asked to juggle knives.
Stupid, he said. Dangerous.
Charlene agreed. Said she'd seen somebody in Phoenix pass out from smoke in a pallet shelter. Gail heard them.
She went back to the McDonald's and spent two hours reading about carbon monoxide poisoning from Korean and systems. And it scared her. Tens of thousands of people sick in the 1960s and 70s. Thousands dead. Always the same cause. Cracks in the floor, gas seeping up while people slept. But the deaths happened when Koreans switched from wood to coal briquettes, coal smolders, and produces far more carbon monoxide. The cracks formed in floors heated and cooled thousands of times over decades.
And the rooms were sealed shut with no fresh air. Gail was planning to burn dry wood in an open pit, run the smoke through a sealed trench, vent it through a chimney pipe. The smoke would never enter the shelter. If it leaked, she'd smell it. She tried explaining this to Dale. Dale said it was still stupid, so she stopped talking about it and started building. Late October 2023, the weather was still in the 50s during the day, low30s at night. She had time. Not a lot, but some. The cinder blocks were the easy part. The pile behind the warehouse had maybe 150 of them, plus a bunch of broken halves. They'd been sitting there since the roofing company closed. Nobody wanted them. She carried six at a time in a milk crate. She'd found four trips a day over the course of a week. She stacked them near the south side of the warehouse wall where the wall gave wind protection and where the ground was flat and firm, packed gravel and calish. She picked a footprint 8 ft x 6 ft about the size of a walk-in closet, which when you think about it is bigger than a tent and smaller than a shed and roughly the size of a space that one person's body heat can almost warm on its own. Almost. Not quite, but almost. The first thing she dug was the trench. This was the hard part, the actual digging. She had a garden spade she'd found at a yard sale box marked free on Edith Boulevard and a flathead screwdriver she used for prying and her hands. The ground in that part of Albuquerque is not soft. It's kish over clay over rock. And she was digging a trench 12 in deep and 10 in wide and about 7 ft long. Starting from a spot 6 feet outside her planned footprint, running under where the floor would be, and ending at the far wall where the chimney pipe would go up. She dug for 4 days, 2 or 3 hours a day, usually in the afternoon, when the sun was warm enough that the physical work felt good instead of punishing. The blisters on her palms opened and closed and opened again. The scar on her left hand from the box cutter split along its length and bled into the dirt, and she wrapped it with a strip of t-shirt and kept going. By the time the trench was done, her shoulders achd in a way that reminded her of the rotator cuff injury, the dull, thick ache under the socket, and she sat next to the trench one evening and thought seriously about whether this was insane, whether Dale was right, whether she was building an elaborate way to poison herself because she'd read some articles on her phone and gotten excited. She sat there for a long time. The trench was a wound in the ground, 7 ft of raw dirt, and it didn't look like anything yet. It looked like someone had started digging a grave and gotten bored. But the parking lot thing kept coming back to her. The parking lot at the Walmart on Kors Boulevard, the one she'd walked across in July. The heat so intense through her shoes she could feel it in her ankles. That heat was real. It came from mass. From the weight of material that had absorbed energy and was giving it back. That was just physics, not a trick, not something she'd invented.
just the world doing what the world does. She went to the McDonald's that night and looked up one more thing. She looked up how deep you have to go before the ground temperature stops changing with the weather. The answer for Albuquerque was about 4 to 5 ft. At that depth, the ground sits at roughly 55° year round. She wasn't digging that deep. Her trench was only 12 in down.
But even 12 in of packed dirt over a heat channel meant the heat had somewhere to go that wasn't straight up into the air. It meant the dirt was a sponge and the smoke was the water and the sponge would hold the water after the faucet turned off. That was her analogy. She used it with everyone.
Sponge. Water. Faucet. Nobody cared. She built the walls in early November.
Mortar was the one thing she bought. Two bags of quickrit from Home Depot on Lomus, $11 each, paid for with plasma money from Biolife on Manal. She mixed it in a 5gallon bucket with a stick. Her joints were ugly, thick in some places, non-existent in others. And the first three courses looked like a child's art project. By the fourth, she'd figured out the trick. Less mortar, more tapping. Five courses high, 40 in, just over 3 ft. Not standing height, sitting, kneeling, lying down height. She left a 20-in opening on the south side for a tarp door. The trench ran under the floor from west to east. At the west end outside the shelter was the fire pit, 18 in deep and 2 ft across, lined with broken cinder block halves and red brick chunks from the drainage ditch. At the east end, the trench exited through the bottom of the east wall to a vertical chimney pipe. The pipe was 8 ft of 4-in galvanized steel duct from a drier vent kit she'd found in a dumpster behind a house being renovated on Lead Avenue. 4 in was narrow, probably too narrow. She secured it to the outside wall with wire and zip ties. The roof was corrugated metal from the warehouse pile laid across the walls and waited with cinder blocks on the corners. Now the part that mattered, the trench cover. She had to cover the trench so it became a closed channel. Smoke goes in at the fire pit end, travels through the covered trench under the floor, and exits at the chimney end. The cover has to be solid enough to walk on, tight enough that smoke doesn't leak through, and dense enough to absorb and radiate heat. She used concrete pavers, the 16-in square ones, the kind people use for garden paths, a construction site on Broadway, and Cole was throwing them out, cracked and stained. And she asked the foreman if she could take them, and he said, "Go ahead." She carried them in the milk crate, four at a time, a mile and a half each way, over 3 days, 16 pavers. They covered the trench with about an inch of overlap on each side, resting on the lips of dirt she'd left along the trench edges. The gaps between pavers mattered.
She needed those gaps sealed. She mixed a thin slurry of quicret and water, thinner than the wall mortar, almost like a paste, and she spread it over every joint between the pavers and pressed it into the cracks with her fingers, and then smoothed it with the back of the spade. She did this twice.
Let the first coat dry overnight, then another thin layer the next morning. On top of the sealed pavers, she laid a piece of vinyl flooring she'd cut from a roll in the warehouse. The roll had been sitting in the loading dock for 4 years, still in its plastic wrap, warped on the edges, but fine in the middle. She cut a piece 6 ft x 4 ft with a utility knife, and laid it directly on the pavers. The vinyl was her gas barrier. Even if the concrete slurry cracked from heating and cooling, the vinyl would stretch where the crack opened. Gas doesn't pass through vinyl. This was not something she'd read. This was something she'd figured out, sitting on the ground next to her half-built shelter with a square of vinyl in her hand, bending it, noticing it didn't tear, noticing it didn't have holes, thinking about what happens to a gas that encounters a solid sheet with no holes in it. It stops.
That's what happens. It stops. She told Dale about the vinyl. He said vinyl melts. She said the floor wouldn't get hot enough to melt vinyl. He asked how she knew. She said she'd looked it up.
Vinyl flooring melts at about 300°. The floor of a Korean and Doll house, the kind people have been sitting on for centuries, maxes out around 80 or 90° F.
her floor would be somewhere between 80 and 100° if everything worked the way the internet said it would. Dale said the internet said a lot of things. She finished the build on November 16th, 2023. The total cost, if you added it up, was $22 for the mortar. Everything else was scavenged or free. She'd spent roughly 3 weeks on it, working 2 to four hours a day alone. The first test fire was that evening. She gathered a small pile of dry wood. Not much. A few broken pallet slats and some branches from the cottonwood trees along the drainage ditch snapped to 12-in lengths. She arranged them in the fire pit, crumpled newspaper underneath, and lit it with a dollar tree lighter. The fire caught, smoke rose, and for the first 30 seconds or so, the smoke just went straight up from the fire pit into the sky, same as any fire. And Gail stood there watching it and thinking, "This is not going to work. This was three weeks wasted." Dale is going to say, "I told you so." And he'll be right. Then something shifted.
The chimney pipe on the far side of the shelter began to warm up. She couldn't see it happen, but she could hear a faint hissing sound, like air being pulled through a tight space, and the smoke above the fire pit started to lean, not blow, because there was no wind. Lean. tilt. The column of smoke that had been going straight up began curving toward the trench opening at the base of the fire pit. Like the smoke was being sucked sideways into the ground.
Within 2 minutes, most of the smoke was gone from above the fire. It was entering the trench, traveling under the floor of the shelter and coming out the chimney pipe on the other side. She walked around to the east wall and looked at the pipe. A thin gray wisp was rising from the top. She put her hand near the pipe. Warm. She went inside the shelter and put her hand on the floor.
Warm. Not hot. Not even particularly warm, but warmer than the dirt outside.
Warmer than cinder block that hadn't been heated. The pavers over the trench had a temperature her palm could detect, and it was different from the temperature of everything else, and it was rising. She sat on the floor and waited. The fire burned for about 2 hours. She fed it slowly, small pieces, keeping the flame active, but not huge.
She didn't want a bonfire. She wanted a sustained, moderate fire that would send a steady stream of hot smoke through the channel for as long as possible. The Korean and article said a few hours of fire could heat a floor for 8 to 12 hours. She wanted to test that. At about 9:00 p.m., she let the fire die, stopped adding wood, the flames dropped, the coals glowed, and within 20 minutes, the fire pit held nothing but gray ash and a few orange embers that winked and faded.
She went inside and closed the tarp door. The floor was warm, not like a heated car seat, not like a hot parking lot, but warm the way sunbaked concrete is warm at 7:00 p.m. A gentle broad warmth coming up through the vinyl, through the pavers, from the earth underneath where the smoke had traveled and deposited its heat into the dirt and the stone and the concrete. She lay down on a sleeping bag on the floor, and the warmth came through the bag and into her back and her shoulders and the backs of her legs. And it was the first time in 14 months that heat reached her from below, from the ground, from the surface she was lying on instead of from a fire she had to face. She checked the meat thermometer at 10 p.m. It read 58° inside the shelter. Outside, according to the Weather Channel app on her phone, it was 34. She checked it again at midnight, 61° inside. The floor was releasing heat. The temperature was going up even though the fire had been out for 3 hours. She fell asleep at 7:00 a.m. February light coming through the gap between the tarp door and the cinder block wall. She woke up and tapped the thermometer. The needle stuck. She tapped it again and it settled on 65.
Outside, according to her phone, it was 29°. The fire had been out for 9, maybe 10 hours. 65° with no fuel burning anywhere. She put her hand on the floor, still warm, noticeably warm. The pavers over the trench channel were warmer than the pavers to the sides. The center of the floor, directly over the path the smoke had taken, was the warmest spot.
The edges were cooler, but still above ambient. The heat was radiating up from the earth, from the mass of dirt and stone and concrete that had soaked up the smoke's energy for 2 hours, and was now spending it back. slowly, the way a sponge releases water when you set it on a counter and walk away. She walked to Dale's spot by the fire pit. Dale was sitting in a camp chair, wearing a sleeping bag like a cape, drinking instant coffee heated on a propane burner. It was 29° and he looked miserable. She said 65°.
Dale said, "Bullshit." She said, "Come look." He didn't go look. Not that morning. But Charlene did 3 days later after Gail ran the fire again and got similar results. Charlene stood inside the shelter and looked at the thermometer and looked at the floor and looked at Gail and said, "What the hell?" Gail showed her the trench, lifted the edge of the vinyl flooring and pointed to the pa joints underneath, sealed with quickret slurry. Showed her the fire pit outside, now just a circle of blackened cinder block halves with cold ash at the bottom. showed her the chimney pipe, which was cold, too. No smoke, no heat, nothing. And then she pointed to the floor. The floor was still warm. Charlene asked how. And Gail tried to explain. And this is where it gets interesting because Gail's explanation was not a textbook explanation. It wasn't physics. It was the parking lot. You know how when you walk across a parking lot at night in the summer, she said, and the asphalt is still hot even though the sun went down an hour ago, the asphalt weighs a lot, it soaked up the sun all day. Now it's giving the heat back. My floor does the same thing. The smoke is the sun. The dirt under the pavers is the asphalt.
The smoke heats the dirt, and the dirt holds the heat because dirt is heavy, and heavy things hold heat. And the fire is outside. Gail said the smoke never enters the room. It goes through the trench, under the floor, and out the pipe on the other end. The pipe is the chimney. The chimney pulls the smoke through because hot air rises. And when hot air rises inside the pipe, it sucks new air through the trench. And the new air brings new heat from the fire. And the heat dumps into the dirt on the way through. When the fire dies, the smoke stops, but the heat is already in the ground. It's in the pavers and the dirt and the cinder block edges of the trench. And those things are heavy and dense and slow to cool and they radiate that heat up through the floor for hours. Charlene said, "And you don't breathe any smoke?" Gail said, "Smell the air in here." Charlene smelled it.
Dirt, old vinyl, a faint mineral smell from the cinder blocks. No smoke, no char. No combustion products of any kind. The air inside Gail's shelter smelled like a basement, not like a campfire. Gail had also bought a carbon monoxide detector, $9 at Walmart, battery powered. She'd velcroed it to the inside of the east wall, 6 in above the floor, right near where the trench exited to the chimney pipe. The display read zero. It had read zero every time she'd checked it, during a fire and after for two weeks of testing. The thing about carbon monoxide is that it's a failure product. It comes from incomplete combustion, from smoldering, from coal and charcoal in damp wood burning at low temperatures. Gale burned dry wood at high temperatures in an open pit with plenty of air. The combustion was close to complete. The smoke was mostly carbon dioxide and water vapor and particulate matter, and all of it traveled through a sealed channel and out of pipe without entering the shelter at all. The trench was sealed. The paver joints were sealed. The vinyl floor was an unbroken gas barrier. The chimney maintained suction. The system was not airtight because no improvised system is, but it was tight enough that the pressure gradient worked in her favor.
The chimney pulled air through the system from fire pit to chimney pipe, which meant any air leakage around the pavers would be inward, pulling room air into the trench rather than pushing trench air into the room. The chimney's pull was her safety valve, and she hadn't fully articulated this at first, but she understood it intuitively after the first time she lit the fire and watched the smoke get sucked into the ground. Word got around the camp slowly.
Charlene told a woman named Grace, who slept in one of the tents. Grace told her partner, Rick. Rick came and stood in Gail's doorway one evening and said, "I don't get it. Where's the heater?"
Gail pointed at the floor. The floor is the heater," she said. Rick looked at her like she was speaking another language. This was the reaction she got most often, not hostility. Not after the first few weeks when the thing clearly worked and didn't kill her. Confusion.
People couldn't process the idea because it didn't match any heating system they'd encountered. A heater is a thing.
A space heater, a propane burner, a wood stove, a furnace. It's an object that gets hot and you sit near it. Gail's system had no object. The floor was the heater, and the floor just looked like a floor. And the fire was outside where you couldn't even see it from the shelter. And the chimney was just a pipe on the wall. "And where was the heater?"
She started calling it the sponge. "My sponge," she'd say. "I'm charging the sponge tonight." She meant she was running the fire for a couple of hours to put heat into the ground. After 2 hours of fire, the sponge would stay warm for 8 to 12 hours, depending on how cold it was outside. On nights in the 40s, one fire could last all night and into the next afternoon. On nights in the teens, which happened a few times in January 2024, she'd need to run a second fire early in the morning to keep the floor above 55. But even 55° is a different universe than 20°. 55° means you can sleep in a regular sleeping bag and not shiver. It means your water bottle doesn't freeze. It means you can take off your boots and feel your feet.
It means you can sit and think and not spend every bit of your mental energy on the single problem of being cold. Dale came around in December not because he believed in the system because his tent had a broken zipper and it was supposed to drop into the 20s overnight. And Gail said he could sleep on the warm side of her floor if he wanted. He went in skeptical and came out at 6:00 a.m. with a strange look on his face. He said the floor was warm. He said he hadn't done that since August. He still called it stupid, but he started sleeping there on the coldest nights. And he stopped saying she was going to kill herself with gas. By January 2024, Gail had made a few improvements. She'd found a second piece of chimney pipe, extending the stack to about 10 ft above ground level, which improved the draft.
She'd packed dirt against the outside of the cinder block walls on three sides, burming them about 18 in up, which added mass and cut wind infiltration. She'd sealed the gap between the corrugated metal roof and the top of the walls with strips of foam pool noodles she'd cut lengthwise, stuffing them into the cracks. The shelter wasn't airtight, and she didn't want it to be, but it was tighter. And she'd added a second trench. This was the real upgrade. The original trench ran straight through the middle of the floor, west to east. The second trench branched off the first one about 2 ft into the shelter and ran along the south wall parallel to the first before reconnecting at the east wall near the chimney pipe. A U-shape, more or less, two channels of hot smoke under the floor instead of one. She dug the second trench from inside the shelter on her knees with the spade and a soup can she'd cut the bottom out of and used as a scoop. It took her 3 days because she was working in a confined space. She sealed it the same way.
Pavers and slurry and vinyl. The second channel doubled the amount of floor that got warm. Now almost the entire floor was over a smoke channel, and the heat distribution was more even. January 14th, 2024. The temperature in Albuquerque dropped to 12° at 4:00 a.m.
The National Weather Service had issued a hard freeze warning. Gail ran her fire for 3 hours that evening from about 6:00 to 9:00 p.m. burning dry pallet wood and cottonwood branches. She let it die and went inside. At midnight, the floor was 62°.
At 4:00 a.m., when the air outside hit 12, the floor was 57 and the air inside the shelter was 53. At 7:00 a.m., when the outside temperature climbed to 19, the inside was back up to 56. She hadn't run the fire again. She hadn't done anything. The dirt under her floor was still radiating the heat from yesterday's fire, and it was enough to keep the shelter 40° warmer than outside. For comparison, Dale in his repaired tent said it was so cold that night he couldn't feel his fingers when he woke up and had to run warm water over them from a jug. Charlene in the hot tub shell said she didn't sleep at all. Grace and Rick ran their propane heater inside their tent, which kept them warm, but cost $3 in fuel and left them both with headaches from the fumes.
Gail spent nothing, burned nothing that night. 40° of warmth stored in dirt left over from a fire she'd lit the previous evening. There's a word for what Gail built. The word is and it's Korean. It means warm stone. Koreans have been heating their homes this way for at least a thousand years, maybe longer.
The oldest known andal dates to about 1,000 BC, found in archaeological sites in northern Korea. The Chinese have their own version called a Kang, which is a heated sleeping platform. The Romans had a version called a hypocost, channels under the floors of bathous carrying hot air from a furnace outside.
Gail didn't know any of this when she started digging. She learned some of it along the way, reading at the McDonald's, piecing together articles and forums and YouTube videos about rocket mass heaters and off-grid heating and Korean architecture. She learned that the principle she'd stumbled onto was old. Really old. Older than Christianity, older than written European history. People had been putting fire underground and sleeping on the warm dirt above it since before the Bronze Age. But she hadn't stumbled onto it through history. She'd stumbled onto it through a parking lot, through the feeling of hot asphalt under thin shoes on a July evening in Albuquerque.
Through the observation that heavy things hold heat and light things lose it, and that the ground under your body is the heaviest thing you'll ever touch.
That's the part that's hard to get your head around, even now. The idea is simple. Embarrassingly simple. Fire goes under the ground. Ground gets warm.
ground stays warm. You lie on the warm ground. Children can understand this.
Children do understand this in Korea where heated floors are so normal that kids grow up sitting on the floor because the floor is the most comfortable spot in the house. But in Albuquerque, New Mexico, in a camp behind a dead roofing warehouse in the winter of 2023 and 2024, nobody had heard of it. Nobody had thought of it.
Gail thought of it because she got annoyed at a campfire and typed a question into Google at 2 in the morning. And then she spent three weeks digging a trench and stacking cinder blocks and then she was warm and the peers who called her stupid were cold.
The system had problems. She's the first to say so. The chimney pipe was too narrow and clogged with creassote twice, once in late November and once in February. and both times she had to take it apart and scrape the inside with a wire brush she'd made from a coat hanger. The 4-in pipe restricted air flow, which meant the draft was marginal on calm nights, and if she burned anything other than very dry wood, the fire would smolder instead of flaming, and the smoke would back up, and some of it would seep around the pa joints, and she'd smell it inside and have to open the tarp door until the chimney caught again. She dealt with the creassote problem by burning hotter. She figured this out the same way she'd figured out everything else, by noticing a pattern in reasoning from it. When the fire burned hot with plenty of air and dry wood, the chimney didn't clog. When the fire smoldered, it clogged fast. The reason, which she read about later, is that creassote condenses when smoke cools too much before exiting. Hot fires make hot smoke that stays gaseous all the way through the chimney. Cool fires make cool smoke that condenses into tar on the pipe walls. So she kept the fire hot, small and hot, not big and smoky. A few pieces of dry pallet wood at a time, broken small, stacked loose so air could circulate through the pile. The fire crackled and flamed and sent clean, near invisible exhaust through the trench and out the pipe. The creassote problem mostly went away. The other problem was the roof. Corrugated metal with pool ludal insulation is basically nothing.
Heat rises. The warm floor heated the air in the shelter and the warm air rose to the ceiling and the metal ceiling conducted that heat straight outside.
She was losing a huge percentage of her stored heat through the roof. In January, she laid a double layer of cardboard across the underside of the metal, held in place with duct tape and the weight of the blocks at the edges.
Cardboard is a terrible insulation material in the rain, but inside a dry shelter under a metal roof, it's decent.
The cardboard trapped a dead air gap between itself and the metal, and dead air is a decent insulator. The overnight temperature in the shelter jumped about 4° after she added the cardboard. She also thought about insulating the walls, but the earth burming she' done on three sides was already doing most of that work. The dirt pressed against the outside of the cinder blocks was adding mass and blocking wind, and the cinder blocks themselves, while poor insulators were storing some heat from the warm interior air and releasing it slowly.
The south wall, the one with the door, was the weak point, and she addressed it by hanging a second tarp over the opening with a 4-in gap between the two tarps, creating another dead air buffer.
February 2024 was mild in Albuquerque, highs in the 50s, lows in the upper 20s, and low30s. Gail's shelter held between 58 and 67° most nights with a single 2-hour fire in the evening. She stopped wearing her coat inside. She stopped wearing a hat to bed. She slept in a t-shirt and sweatpants on a sleeping bag she unzipped and used as a blanket on a floor that was warm to the touch in a structure that cost $22 and 3 weeks of labor and $0 per night to operate. Her fuel cost was approximately nothing.
Palletwood is free. Cottonwood branches are free. Drywood is everywhere in Albuquerque if you know where to look.
and Gail looked and she burned two to three arm loads per evening and that was enough to charge the sponge for the night. A man from the camp across the drainage ditch, a guy named Earl, who Gail had never spoken to before, came over in late February after hearing about the shelter from somebody who'd heard about it from somebody. He stood inside and looked at the thermometer and looked at the floor and looked at the chimney pipe and looked at Gail and said, "You should patent this." Gail said it was a thousand years old. Nobody could patent it. She didn't invent it.
The Koreans invented it. She just built one in a vacant lot. Earl said he didn't care who invented it. He wanted one.
Could she build one for him? She said she'd show him how, but he had to dig the trench himself. By March 2024, two other people in the camp had started digging. Charlene abandoned the hot tub shell and began collecting cinder blocks. Rick, who was handy with tools, found a longer section of 6-in stove pipe at a construction dumpster on Carlile and used it for his chimney. And his draft was better than Gail's from day one because the wider pipe let more air move. Dale never built one. Dale said he'd rather freeze than dig a 7-ft trench in Kish. But Dale slept on Gail's floor about two nights a week all winter, and he brought wood sometimes.
And he stopped calling it stupid. He called it the oven. Gail's oven, which bothered Gail because it wasn't an oven.
An oven heats from all around. And her system heated from below, which is fundamentally different, and she would argue better. Heat from below rises through the room. Heat from a space heater or a fire pit goes mostly sideways and up, and the floor stays cold, and your feet are always the coldest part of you, and the warmest air is at the ceiling, where it does nobody any good. Her floor was the warmest surface in the room. Heat rose from it through the sleeping bag, through the air, toward the ceiling. The temperature at floor level was always higher than the temperature at head level, which is the exact opposite of how most heated rooms work. In a normal room with a furnace or a heater, the air at the ceiling might be 75 and the air at the floor might be 60. In Gail's shelter, the floor was 68 and the air at head height was 60. You were warmer lying down than standing up. The room was designed without her fully intending it for sleeping, for resting, for the thing a homeless person needs most and gets leased. The camp was cleared in April 2024. City crews came through and posted 72-hour notices and then came back with trucks and took everything. Gail's shelter was torn down. The cinder blocks were loaded into a dump truck. The pavers were scattered. The chimney pipe was bent and tossed into the drainage ditch. She watched from the parking lot of the Autozone across Gibson Boulevard.
She didn't fight it. She had known it was coming. The notices had been posted for weeks. She'd taken the meat thermometer and the carbon monoxide detector and her sleeping bag in a plastic bin of clothes and walked away before the crews arrived. She told me when I talked to her in June 2024 at the Albuquerque Opportunity Center on Second Street that she could build another one in 3 weeks. Two, if she had help. The materials are in every vacant lot and every dumpster in every city in America.
Cinder blocks, pavers, mortar, pipe, vinyl, $30. Maybe the knowledge is on your phone. The Koreans figured this out before they had phones or mortar or cinder blocks or galvanized pipe. They figured it out with flat stones and clay and a hole in the ground and a thousand winters of paying attention to where the heat goes when you stop watching it.
Gail figured it out with a cracked phone and a parking lot and a question she couldn't stop asking. Where does the heat go? Into the ground? Into the stone? Into the heavy, dense, patient mass of the earth under your body where it waits? where it holds, where it stays long after the fire has forgotten it was burning. That's where the heat goes. And 65° on a February morning with the fire out and the ash cold and the sky still dark over the Sandia Mountains is what it feels like when you finally know the
Related Videos
Is dark matter real? - Why can't we find it? - physicist explains | Don Lincoln and Lex Fridman
LexClips
1K views•2026-05-30
Saptarshi Basu - Spectacular Voyage of Droplets: A Multiscale Journey to Extreme Flow Conditions
DAlembert-SU-CNRS
152 views•2026-06-02
A 6.0 Just Hit Hawaii — And It Came From The Wrong Place
TerraWatchHQ
115 views•2026-06-03
The Split-Second Mistake That Made Bouncing Bettys So Deadly
NoMansLandChannel
253 views•2026-06-02
Nobody Expected This Lava Reaction 🤯 #faits #facts
TendzDora
28K views•2026-05-30
The Difference In Charged And Neutral Particles
heavybrainspace
959 views•2026-05-29
The Silent Memory of Glass
UnchartedScienceworld
146 views•2026-05-30
A380 vs Every Vehicles Crash Test Challenge | Which One Win?
BeamLap
163 views•2026-05-29











