#5454: Why a 4-Story-Deep Garage Beats a 20th-Floor Concrete Core

A deep garage should be cool. It isn't. A 20th-floor concrete core should be neutral. It isn't. Here's the physics.

Featuring
Listen
0:00
0:00
Episode Details
Episode ID
MWP-5637
Published
Duration
23:58
Audio
Direct link
Pipeline
V5.2
TTS Engine
chatterbox-regular
Script Writing Agent
DeepSeek 4.1 Flash

AI-Generated Content: This podcast is created using AI personas. Please verify any important information independently.

The geothermal gradient runs about 25 to 30 degrees Celsius per kilometer of depth in continental crust. A parking garage four floors down is 12 to 15 meters. That works out to roughly four tenths of a degree of warming from the Earth's interior—less than half a degree. To get even one degree of core-driven heat you'd need to go 35 or 40 meters down. The core is 6,000 kilometers away; four floors is a rounding error on a rounding error.

So why was that garage so hot? Because weather and climate only penetrate the top 10 to 20 meters of ground. Below that, soil sits at the mean annual ground temperature—the year-round average air temperature, frozen into the earth. In a hot, arid climate that average can be 25 to 30 degrees Celsius. The ground isn't cool; it's the local average. Then add cars: a garage full of recently driven vehicles is a room full of radiators, with hot engines, exhaust, brakes, and tires, and typically no mechanical cooling. Concrete does the rest, storing enormous thermal energy per degree and re-radiating it slowly. Depth makes it worse, not better—the surrounding earth insulates the space from nighttime cooling and wind, so heat accumulates instead of bleeding off.

The twentieth-floor concrete core is the same liability in a different configuration. It never sees direct sun, but it's structurally continuous with exterior columns, floor slabs, and facade. Solar radiation heats the south and west faces and the roof, and that heat conducts through the concrete into the interior—a heat highway from the sunlit skin to the middle of the building. The stack effect adds rising warm air through elevator shafts, stairwells, and utility risers. And the core is surrounded by apartments full of people, cooking, electronics, and hot water pipes, with mechanical and electrical risers running through it. As MIT's Christoph Reinhart notes, thermal mass is a liability when a space is used intermittently and heat flow through the envelope never reverses. The mass isn't moderating anything—it's a sponge that fills up and stays full.

Downloads

Episode Audio

Download the full episode as an MP3 file

Download MP3
Transcript (TXT)

Plain text transcript file

Transcript (PDF)

Formatted PDF with styling

#5454: Why a 4-Story-Deep Garage Beats a 20th-Floor Concrete Core

Corn
Herman, quick question. If I put you in a basement in Jerusalem in August, what temperature are you expecting?
Herman
Cooler than upstairs. That's the whole point of a basement.
Corn
Right. And if I put you four floors underground, in a parking garage, in a hot country, in the middle of a war?
Herman
Then I'd expect to be wrong, apparently.
Corn
Hannah wrote in this week. Not Daniel, Hannah.
Herman
Oh, good. She's the one who actually knows buildings.
Corn
She does. And she's got two observations that don't fit together. The first one is from the Iran war. They were spending time in a parking garage about four floors underground, and it was unbelievably hot. Which felt backwards, because basements are supposed to be cool. Daniel told her, well, of course, you're closer to the Earth's core.
Herman
Daniel said that? Daniel, who writes our prompts about agentic AI and workflow orchestration.
Corn
Same Daniel. And Hannah refuses to believe the core could matter over the distance of four floors. She's right to refuse.
Herman
She's extremely right to refuse.
Corn
Then the second observation. They live high up in a tower now, and the concrete core of the building around the twentieth floor roasts in summer. Buried inside the building, surrounded by apartments, several floors above it, no direct sunlight. Their own apartment getting hot from solar radiation, she understands. The protected interior concrete core she does not understand.
Herman
That one's interesting.
Corn
And then the actual question. What determines temperature at different heights and depths in a building? Thermal mass, ventilation, trapped air, mechanical systems, heat rising, the surrounding ground, or something else. And how much does elevation itself actually matter.
Herman
That's a good prompt.
Corn
It's a very good prompt. So let's start by killing the Earth's core theory with arithmetic, and then figure out what's actually going on.
Herman
Let's do it. And the arithmetic really is brutal. The geothermal gradient, the rate at which the Earth gets hotter as you go down, is about twenty-five to thirty degrees Celsius per kilometer in continental crust. That's the number.
Corn
Per kilometer.
Herman
Per kilometer. A four-floor-deep garage is twelve to fifteen meters. So you take thirty degrees per thousand meters, multiply by fifteen meters, and you get about four tenths of a degree. Less than half a degree of warming from the Earth's interior.
Corn
Half a degree.
Herman
To get even one degree of core-driven warming you'd have to go thirty-five, forty meters down. And the core is six thousand kilometers away. Four floors is a rounding error on a rounding error.
Corn
So Daniel's theory is off by roughly two orders of magnitude.
Herman
It's off by more than that, honestly. It's not even the right phenomenon. The heat coming up from the Earth's interior is real, it's just so diffuse at that scale that it's completely swamped by everything else happening in the garage.
Corn
Which means the "basement equals cool" intuition has to be explained some other way. Because it's not wrong, exactly. It's just local.
Herman
It's a temperate-climate intuition. Here's the thing that actually governs it. Weather and climate only penetrate the top ten to twenty meters of ground. Below that, the temperature of the soil is essentially the mean annual ground temperature. The year-round average air temperature at that location, basically frozen into the earth.
Corn
Which is why ground-source heat pumps work.
Herman
Exactly why. You drill down ten, twenty meters and you find a temperature that barely moves all year. That stability is the whole product. It's moderate in a temperate climate. In a hot, arid climate, the mean annual ground temperature can easily be twenty-five to thirty degrees Celsius. So a deep garage in that climate starts warm before a single car drives in.
Corn
The ground isn't cool. The ground is the local average.
Herman
The ground is the local average, and the local average in a hot place is hot. A basement in Ireland is cool. A basement in the Gulf is not.
Corn
Okay. So the ground is warm. That's the floor of the problem. What's the ceiling?
Herman
Cars. Cars are enormous heat sources. A car that has just driven in has a hot engine, hot exhaust, hot brakes, hot tires. A garage full of recently driven cars is a room full of radiators. And then there's no air conditioning, and typically minimal mechanical ventilation. So all of that heat has nowhere to go.
Corn
It just accumulates.
Herman
It accumulates. And then the concrete does the rest. Concrete has very high volumetric heat capacity. It stores a lot of thermal energy per degree of temperature change. So the floors, the walls, the ceilings, all that mass absorbs heat all day and re-radiates it slowly. A garage is a giant concrete box that soaks up heat and holds it.
Corn
And the depth, which is the thing that felt like the explanation, actually makes it worse.
Herman
It makes it worse. Being underground insulates you from night-time cooling and from wind. The surrounding earth is a thermal blanket. A deep garage loses heat far more slowly than a surface structure would. So the heat that builds up during the day doesn't bleed off overnight the way it would above ground.
Corn
So the honest answer to Hannah's first question is that the garage is hot because it's a poorly ventilated concrete box full of hot cars, sitting in ground that's already warm because the local climate is hot. And the depth contributes by trapping heat, not by conducting it up from the core.
Herman
That's the whole thing. Depth is a contributing factor in exactly the opposite direction from what Daniel assumed.
Corn
There's a framing from Christoph Reinhart at MIT that fits here. He works on environmental technologies in buildings, and he makes the point that thermal mass may act as a liability to keeping a space comfortable when the space is only used intermittently. And that thermal mass has really no effect if the direction of heat flow through the building envelope stays constant for extended periods.
Herman
That second clause is the garage exactly. If heat is flowing in and never reversing, the mass isn't moderating anything. It's just a sponge that fills up and stays full.
Corn
A liability, not a buffer.
Herman
A liability. And that's the bridge into the second half of Hannah's question, because the twentieth-floor concrete core is the same liability in a completely different configuration.
Corn
Let's take the garage first, because it's the cleaner case, and the numbers here are almost comically decisive.
Herman
They are. And I want to be fair to Daniel for a second. The intuition that going down means going toward heat isn't stupid. It's true in mines, it's true in deep boreholes, it's true in geothermal wells. It's just not true at fifteen meters.
Corn
It's true at scales where the gradient has room to work.
Herman
Right. At a kilometer down you've gained twenty-five to thirty degrees and you'd absolutely feel it. At fifteen meters you've gained a fraction of a degree and you'd never notice it next to a warm engine block.
Corn
Which is the actual lesson. The core theory isn't wrong about physics. It's wrong about scale.
Herman
It's a scale error. Which is a very common kind of error, honestly. People reach for the dramatic mechanism when the boring one is doing all the work.
Corn
The boring one being ventilation.
Herman
The boring one being ventilation, and the second boring one being thermal mass, and the third boring one being that the local ground is warm.
Corn
So let's put a number on the garage. If the local mean annual ground temperature is, say, twenty-eight degrees, and the garage is full of cars that have been driven in over the course of a day, and there's no mechanical cooling, what are we actually looking at inside?
Herman
Depending on the garage, you can easily be ten, fifteen degrees above the outside air temperature. Some of these structures run into the forties Celsius internally in summer. And the reason it's so striking is that you walk in expecting cave conditions and you get oven conditions.
Corn
Because you've been told your whole life that underground means cool.
Herman
Underground means stable. Stable is only cool if the average is cool. That's the whole correction.
Corn
There's a detail in here I want to pull on. The concrete absorbs heat all day and releases it slowly. So when is the garage actually at its worst?
Herman
Late. Not at noon. The mass lags the input. So the hottest part of the day inside that structure might be late afternoon or into the evening, after the cars have all come in and the concrete has had hours to soak.
Corn
So if you were going to work on that garage, you'd want to be there at three in the morning.
Herman
You'd want to be there at three in the morning, and even then the concrete is still giving back what it took in. That's the thermal lag. It's the same physics that makes a stone building hold heat long after sunset.
Corn
Which is a nice thing, in a house. In a garage, it's a slow cooker.
Herman
It's a slow cooker with a thousand cars in it.
Corn
That's the underground half. Now flip it. Same building, same physics, completely different mechanism, twenty floors up.
Herman
The twentieth-floor core. And this one really is a puzzle, because every naive explanation fails. It's not in the sun. It's surrounded by apartments. There are floors above it. So where is the heat coming from?
Corn
Start with the mass.
Herman
Concrete has very high volumetric heat capacity, which means it takes a lot of energy to change its temperature, and it gives that energy back slowly. The core absorbs heat over the course of the day and releases it hours later. So the core can be at its hottest in the late afternoon or the evening, after peak sun, even though it never sees direct sun.
Corn
It's a battery.
Herman
And the charging current is conduction. The core is not thermally isolated. It's structurally continuous with the exterior columns, the floor slabs, the facade. Solar radiation heats the south and west faces and the roof, and that heat conducts through the concrete into the interior. A continuous concrete structure is a heat highway from the sunlit skin to the middle of the building.
Corn
So the sun is still the source, it's just arriving by a different route.
Herman
It's arriving by conduction instead of by radiation. Which is why you can stand in a windowless core and still be standing in the sun's heat, three hours late.
Corn
That's mechanism two. What's mechanism three?
Herman
The stack effect. Warm air is buoyant and it rises. In a tall building it rises through the elevator shafts, the stairwells, the utility risers, any unsealed penetration in a floor slab. The greater the temperature difference and the taller the structure, the stronger the buoyancy force. So upper floors accumulate heat that has risen from below.
Corn
Heat rising. The one part of folk wisdom that actually holds up.
Herman
It holds up, but not for the reason people think. It's not heat floating up through solid concrete. It's air moving through shafts and gaps, driven by pressure. And in summer the effect actually reverses, but it's weaker. Cool conditioned air sinks down the shafts and exfiltrates at lower levels, while upper floors can still trap rising warm air.
Corn
So either way the upper floors run warm.
Herman
Either way the upper floors run warm. This is a well-documented, measurable phenomenon. Upper floors of high-rises are systematically warmer in summer, and it's not subtle.
Corn
Mechanism four.
Herman
Waste heat from the building itself. Every apartment around that core is a heat source. People, cooking, electronics, hot water pipes. And especially the mechanical and electrical risers, which often run through or right next to the core. Elevator machinery, electrical risers, plumbing stacks. A core surrounded by apartments is surrounded by dozens of small heaters.
Corn
And mechanism five is the one that ties it back to the garage.
Herman
No ventilation. The core is by definition the most enclosed, least ventilated part of the building. Heat that reaches it cannot escape. It's the same trapped-heat problem as the garage, just at the other end of the building.
Corn
And six?
Herman
Radiant and conducted heat from the surrounding apartments. Even with no direct sun, the core is being heated from every direction by the warm apartments around it. In summer, when all of those apartments are hot, the core is sitting in the middle of a warm ring.
Corn
So the twentieth-floor core is hot because it's a massive concrete heat battery that absorbs conducted heat from the sunlit structure, receives waste heat from the surrounding apartments and the risers, sits at the top of a stack-effect chimney pushing heat upward, and has no ventilation to shed any of it.
Herman
That's the full list. And notice that not one item on it is "it's high up, so it's closer to something."
Corn
Which brings us to the actual elevation question. How much does height itself matter?
Herman
Almost nothing, in the sense people mean. A twenty-floor difference is about sixty meters. The geothermal gradient over that distance is roughly one and a half to one point eight degrees, and it's downward, not upward. Going up takes you farther from the core. So if anything the core-heat argument works against the observation.
Corn
The tower should be cooler, by that logic.
Herman
The tower should be cooler by that logic, and it isn't. The atmospheric lapse rate is about six and a half degrees per kilometer, which over sixty meters gives you about four tenths of a degree. Also irrelevant. Gravity, air pressure, proximity to space, all of it, negligible at building scale.
Corn
So what does elevation actually do?
Herman
It determines your position in the stack-effect pressure profile. There's a neutral pressure level in every tall building. Above it, floors are under positive pressure and warm air tends to accumulate and exfiltrate. Below it, floors are under negative pressure and draw in outside air. That's the real vertical variable. It's not altitude. It's where you sit in the chimney.
Corn
You're not higher in the atmosphere. You're higher in the shaft.
Herman
That's the whole thing.
Corn
And there's empirical work confirming this, right? That different floors of the same building run at different temperatures?
Herman
There is. A group out of Switzerland, Sulzer and colleagues, published in Sensors in twenty twenty-two, and what they showed is that a network of thermal comfort sensors can detect differences in comfort at different workplaces within the same building. They could capture spatial and temporal differences across floors over the course of a day. So it's not a theoretical curiosity. You can measure it.
Corn
And there's a twenty twenty-five study on high-rise residential buildings?
Herman
Chen and colleagues in Scientific Reports. They found that solar heat gain leads to heat retention and rising indoor temperature, while ventilation promotes cooling, and that these two forces operate in opposing directions. Which is the whole episode in one sentence. Heat in versus heat out.
Corn
And the trapped-heat theme shows up in the office building literature too.
Herman
It does. Yau and colleagues in twenty eighteen looked at poor ventilation and air mixing in high-rise office buildings, and the finding is what you'd expect. When air doesn't mix and doesn't move, heat pools.
Corn
So the unifying principle is that temperature is a balance of heat in versus heat out. Elevation and depth matter only insofar as they change that balance. Up, through the stack effect. Down, through insulation from night cooling.
Herman
And the Earth's core is never the answer. Not at four floors, not at twenty.
Corn
There's something I want to sit with for a second. Both of Hannah's observations are counterintuitive in opposite directions. The garage is hot when we expect cold. The core is hot when we expect... nothing. Neutral. Inert.
Herman
The core one is stranger, honestly. The garage at least has cars in it. You can point at the cars. The core has no obvious source. It's just a column of concrete in the middle of a building, and it's radiating heat like it's been sitting in the sun.
Corn
Which it has. Just not directly.
Herman
Which it has, by conduction, and by the stack, and by the neighbors.
Corn
I keep coming back to the fact that the concrete is the same material in both cases. In the garage it's a sponge. In the core it's a battery. Same heat capacity, same lag, completely different role in the story.
Herman
The material doesn't determine the behavior. The configuration does. That's true of a lot of building physics, actually. You can't evaluate a material in isolation. You have to ask what it's connected to and what's flowing through it.
Corn
So the concrete core is doing exactly what the garage slab is doing. Storing heat and giving it back late. The difference is where the heat comes from and whether anything is removing it.
Herman
And nothing is removing it in either case. That's the shared failure. No ventilation.
Corn
Twenty floors up, four floors down, and the answer is the same. Nothing is moving the air.
Herman
The garage has no mechanical ventilation and the core has no windows and no dedicated exhaust. Both of them are just sitting there, absorbing.
Corn
Let me push on one thing. Hannah says the core roasts in summer specifically. Why summer and not winter?
Herman
Because in winter the direction of heat flow reverses. In winter the core is warmer than the outside air, so heat flows out through the structure, and the building wants to shed heat anyway. In summer the flow reverses and the sunlit skin is hot, so the conduction path runs inward. The core becomes the destination instead of the source.
Corn
And Reinhart's point about thermal mass having no effect when the direction of heat flow stays constant. In summer, the direction stays constant for weeks.
Herman
For weeks. So the mass never gets to do its moderating trick. It just fills up and stays full. Same as the garage.
Corn
There's a nice symmetry there. The garage is a summer problem because the ground is warm and the cars are hot. The core is a summer problem because the conduction reverses and the stack effect stacks up.
Herman
And both get worse as summers get hotter. That's the part I find concerning. The stack effect scales with the indoor-outdoor temperature difference. The thermal mass liability scales with how long the heat load persists. Both of those are trending the wrong way.
Corn
So the twentieth-floor core problem is going to become more common, not less.
Herman
As buildings get taller and summers get hotter, yes. The stack effect gets stronger with height. The mass gets more persistent with longer heat waves. This isn't a quirk of one tower in Jerusalem.
Corn
It's a design consequence.
Herman
It's a design consequence that shows up years after the building opens, when the tenants start complaining about a corridor that's inexplicably warm.
Corn
Hilbert, you've actually worked in one of these things, haven't you?

Hilbert: Nineteen ninety-seven. I did a summer as the night-shift maintenance man in a downtown parking structure. Eight levels. My job was hosing down the decks at three in the morning.
Corn
Hosing them down.

Hilbert: The concrete was still radiating heat from the day's cars. You'd spray a deck and it would steam. At three in the morning.
Herman
That's the thermal lag.

Hilbert: That's what it was. And I'll tell you where it was worst. The middle decks. The top deck at least gets a breeze. The middle decks are boxed in on all sides and the heat just sits there. You could feel it coming off the hoods of cars that had been parked for six hours. The cars don't cool down, they just stop adding.
Corn
They just stop adding.

Hilbert: That's right. And here's the thing that doesn't fit your story. The lowest level of that garage, the one closest to the water table, was the coolest in summer. Cooler than the middle decks. Cooler than the top.
Herman
Interesting.

Hilbert: I kept an infrared thermometer in my truck. Top deck to bottom deck, the bottom was several degrees cooler.
Corn
And you think that's the water table?

Hilbert: I think it's the water table and I think it's the fact that the bottom level had the only real ventilation fans in the building. The only ones that moved any air. I never worked out which one was doing the work.
Herman
Those are two very different explanations.

Hilbert: They are. All I know is what the thermometer said.
Corn
So the bottom of the garage was the coolest part, which complicates the "deeper is hotter" story we just told.

Hilbert: It complicates it. I don't know what to do with it. I just know the number.
Herman
It's actually a good complication. If the bottom level has the fans, then the ventilation is beating the ground temperature. Which is consistent with everything we said about trapped heat being the dominant variable.

Hilbert: Maybe. I've got a thing at four.
Corn
So the water table might be doing something we didn't account for, which is exactly the kind of thing that makes this topic hard to wrap up neatly.
Herman
It's a good reminder that the model is a model. Ventilation, mass, ground temperature, stack effect. Those are the levers. But which one dominates in a given building is an empirical question, and sometimes the answer is the one you didn't expect.
Corn
If temperature is always a heat-balance problem, how much of building design is actually about managing that balance, versus just adding more air conditioning?
Herman
Less than it should be. It's much easier to specify a bigger chiller than to design a building that doesn't need one. And the chiller works, so nobody goes back and asks why the core was hot in the first place.
Corn
And why do we still reach for "closer to the core" or "heat rises" when the real answer is usually no ventilation and a lot of thermal mass?
Herman
Because the folk explanations are about position, and the real explanations are about flow. Position is intuitive. Flow requires you to trace where the heat actually goes.
Corn
And as summers get hotter and buildings get taller, the stack effect and the thermal-mass liability get worse, not better. Which means the twentieth-floor core problem is going to become more common.
Herman
It's going to become more common. And the fix is usually boring. Move the air.
Corn
If you take one thing from this, take the arithmetic. Four floors down is less than half a degree of core heat. The Earth's core is never the answer.
Herman
The next time a room in your building makes no thermal sense, ask where the heat is coming from and where it's going. If nothing is moving the air, you've probably found your answer.
Corn
This has been My Weird Prompts. Thanks as always to our producer, Hilbert Flumingtop.
Herman
Thanks to Hannah for the prompt. If you've ever noticed a room in your own building that makes no thermal sense, that's exactly the kind of thing this show exists for.
Corn
You can find us at my weird prompts dot com, or email us at show at my weird prompts dot com.
Herman
We'll be back soon.
Corn
See you then.

This episode was generated with AI assistance. Hosts Herman and Corn are AI personalities.