Latest / Logosofia no Dia a Dia / Overpressure Not Airtight: How Mamads Actually Work
Transcript
- [Prompt] custom: Hannah and I had a debate over Shabbat. We were talking about how to attempt to get internet into ou | Hosts: corn, herman [Response] Corn: Daniel and Hannah had a Shabbat debate. The question was whether you can run Ethernet into a mamad — an Israeli safe room — without compromising its CBRN protection. Daniel, being the tech guy, pointed at the wall jack and said, look, there's a hole. A few millimeters of gap. That's unfiltered air. Hannah, the architect, said the door doesn't seal hermetically anyway, so the whole airtight premise is already compromised. And they both walked away thinking the other one had just proved their point. Herman: Which is the best kind of Shabbat debate — the kind where both sides think they won, and neither one has actually checked the engineering standard. Corn: And here's the thing. With regional tensions what they are, and every new apartment in Israel being built with one of these rooms, the question of what a mamad actually protects against — and what it doesn't — has stopped being theoretical. People are spending hours in these rooms. They're working from them, putting kids to bed in them, running Ethernet into them. Understanding where the real vulnerabilities are is suddenly very practical. Herman: The answer to their debate is weirder and more interesting than either of them expected. Because it turns out you can have a room with visible gaps — around the door, through wall jacks, around conduit penetrations — and still get genuine CBRN protection. The thing that makes it work is a principle neither of them was thinking about. Corn: Which we'll get to. But first, let's define what we're actually talking about, because the mamad is a very specific piece of engineering that most people outside Israel have never encountered. Herman: A mamad — m-a-m-a-d, it's an acronym in Hebrew for "merkhav mugan dirati," residential protected space — became mandatory in all new Israeli construction starting in the early nineties. The spec is pretty straightforward on paper. Reinforced concrete walls and ceiling, typically thirty to forty centimeters thick. A single heavy steel door that opens outward — that's important, we'll come back to why. A sealed steel window, small, with a blast shutter. And integrated into one wall, a combined air filtration and ventilation unit. Corn: Which is the conspicuous box with the big circular vent that Daniel mentioned. The one everyone assumes is doing something very serious. Herman: And it is. The standard filtration unit in a mamad is rated for CBRN — chemical, biological, radiological, and nuclear particulates. ChatGPT was correct on that point. It uses a multi-stage filter system, typically a HEPA particulate layer plus an activated carbon layer. The HEPA catches dust, smoke, fallout particles, biological aerosols. The carbon adsorbs certain chemical vapors. Corn: Certain chemical vapors. That's a qualifier worth hanging onto. Herman: It is, and we'll get to what it doesn't catch. But the key thing Daniel and Hannah were wrestling with is this apparent contradiction. The filter is CBRN-rated, which sounds like the room should be a sealed bubble. But Hannah's right — the door doesn't seal hermetically. It's heavy, it's steel, it closes against a reinforced frame with a rubber gasket, but it's not an airtight submarine hatch. And Daniel's right — there are wall jacks for coax, phone lines, Ethernet, and each one represents a conduit that runs through the concrete to the outside. Corn: Two people, two correct observations, two completely opposite conclusions about what that means for safety. The architect sees gaps and says the airtight premise is already broken. The tech guy sees a hole and says that's a contamination path. Herman: They're both wrong, in the most interesting way. Because the engineering principle that makes a mamad work isn't hermetic sealing. It's overpressure. Corn: Which is one of those concepts that sounds counterintuitive until you hear it explained once, and then you can never un-see it. Herman: The filtration unit doesn't just clean the air — it pushes filtered air into the room faster than the room can leak it out. That creates positive pressure inside. Maybe twenty to fifty pascals above ambient. Doesn't sound like much, but it means that at every gap — around the door, through the wall jack, around the window frame — air is flowing outward. Contaminants can't drift in against the current. Corn: The gaps everybody was worried about are actually part of the design. They're the exhaust. Herman: They're the exhaust. And that's the shift in thinking that resolves the whole debate. The question isn't "is there a hole." The question is "which way is the air moving through that hole. Corn: Which reframes the whole episode. A mamad isn't a sealed bunker that keeps bad air out by being perfectly tight. It's a leaky room that keeps bad air out by pushing clean air through the leaks faster than anything can get in. And that tradeoff — leaky enough to be livable, pressurized enough to be protective — is where all the interesting engineering lives. Herman: Let's put numbers on this, because the standard is surprisingly specific. Israeli Standard 4218 requires the mamad to maintain a minimum positive pressure of twenty pascals during filtration. That's the official threshold. To put that in perspective, that's about the pressure difference you'd feel if you held your hand out a car window at maybe six kilometers an hour. It's tiny. Corn: Which is why it doesn't feel like anything when you're inside. You don't notice a breeze. You don't feel your ears pop. Herman: And yet that tiny pressure differential is doing all the work. The research on typical mamad doors shows they achieve roughly ten to twenty air changes per hour at a fifty pascal pressure differential. That's the leakiness of the door alone. Ten to twenty complete room volumes of air leaking out every hour when you push to fifty pascals. That is not a hermetic seal by any stretch of the imagination. Corn: The room is hemorrhaging air, and the standard says that's fine as long as the fan is pushing new filtered air in faster than the leaks are letting it out. Herman: That's exactly the design philosophy. The standard doesn't require zero leakage. It requires positive pressure. Those are fundamentally different engineering goals. A zero-leakage room needs perfect seals, airlocks, pressure doors. A positive-pressure room just needs a fan that's correctly sized relative to the total leakage area. Corn: This is where Daniel's wall jack comes in. Because a coax or Ethernet port is what, a few millimeters across? Herman: Maybe six to eight millimeters for an RJ45 jack opening. At twenty to fifty pascals of overpressure, the outward airflow through a hole that size is negligible in terms of pressure loss — the fan won't even notice it — but it's more than sufficient to prevent anything from drifting inward. The air velocity through that tiny gap, even at low pressure, is high enough that particles and gases can't migrate against it. Corn: The wall jack isn't a vulnerability in the pressure boundary. It's just one more exhaust port. Herman: Here's the part that actually matters. The real question isn't the wall plate on the inside of the room. It's whether the conduit behind it is sealed where it exits the building envelope. If that conduit runs through the concrete wall and opens to the outside air somewhere — say, in a junction box on the exterior — and that exterior opening isn't sealed, then you've got a path for unfiltered air to enter the conduit and potentially get drawn in if the pressure differential ever drops. Corn: Which brings us to the concept engineers call equivalent leakage area. A room full of tiny gaps — around the door, through wall jacks, around window frames — all of those add up to a total effective hole size. And the fan just needs to be sized to overcome that total. Herman: And the standard accounts for this. The filtration unit in a mamad is specified to deliver enough airflow to maintain positive pressure given the expected leakage of a properly constructed room. The gaps aren't a bug. They're factored into the sizing calculation from the start. Corn: It's the same principle as a hospital isolation room for infectious disease, just reversed. Those rooms run negative pressure — air flows inward through every gap so pathogens can't escape. Same fluid dynamics, opposite sign. Herman: In a negative-pressure isolation room, you're protecting the world outside from the patient inside. In a mamad, you're protecting the people inside from the world outside. The physics is identical. You create a pressure gradient, and the air does what you tell it to do. Corn: To close the loop on the Shabbat debate. Daniel was right that there's a hole. Hannah was right that the door isn't hermetic. And they were both wrong that those facts defeat the protection. Herman: The protection doesn't come from being sealed. It comes from being pressurized. The holes are the exhaust. The wall jack is a tiny chimney. And as long as the fan is running, everything is flowing in the right direction. Herman: There's a knock-on effect here that's even more counterintuitive. If the room were truly hermetic — perfectly sealed, zero leakage — you'd have a different problem entirely. The filtration unit pulls air from outside, filters it, pushes it in. If nothing can leak out, the fan is just pressurizing a sealed box. And that fan is consuming oxygen. Corn: Wait, the fan itself consumes oxygen? Herman: Not the fan. But the people inside are breathing. In a perfectly sealed room with a recirculating filter, carbon dioxide builds up. The filtration unit in a standard mamad doesn't scrub CO2. It's not a rebreather system. It brings in outside air, filters it, and the old air leaks out through the gaps. If there are no gaps, the air can't cycle. You get stale air, rising CO2, and eventually you've traded chemical protection for hypoxia. Corn: The leaks are literally keeping you alive. The gaps that look like a design flaw are what allow fresh air to cycle through the room. Herman: They're a safety feature. The room needs to be leaky enough that the ventilation system can actually ventilate. The standard is engineered around this balance — tight enough to hold positive pressure, leaky enough to exchange air. Remove the leaks and you've built a slow suffocation chamber. Corn: Which is a pretty good metaphor for a lot of engineering, actually. The thing that looks like the weakness is often the thing that makes the whole system workable. Herman: This connects directly to why Daniel's question about internet access matters more than he probably realized. Because once you understand that a mamad is designed for habitation — not just for hunkering down for ten minutes — the connectivity question stops being about convenience. Corn: During the May twenty twenty-one Gaza conflict, families were spending two to four hours continuously in these rooms. That's not a rocket-attack sprint. That's an extended stay. Kids doing remote school, parents on work calls, everyone trying to maintain some grip on normal life while the sirens cycle on and off. Herman: The IDF Home Front Command explicitly recommends having a battery-powered radio and a charged phone in the mamad. That's not just for emergency alerts. It's for the psychological reality of being stuck in a concrete box for hours. Internet access means you can check what's happening, reassure family members, keep kids distracted with something that isn't the sound of explosions. Corn: If you've ever tried to entertain a five-year-old in a sealed concrete room for three hours without a screen, you understand that internet access is a survivability feature. Herman: It's mental health infrastructure. The mamad protects your body. Connectivity protects your head. During those extended stays in twenty twenty-one, families reported that internet access was critical — not optional — for work continuity, for school, for basic sanity. Corn: The practical question becomes: how do you retrofit connectivity without compromising the pressure boundary? And this is where the conduit seal matters a lot more than the wall plate. Herman: If you're running Ethernet into a mamad, the cable has to pass through the concrete wall somehow. Typically it comes through a conduit — a plastic or metal pipe embedded in the wall. The vulnerability isn't the jack on the interior wall plate. It's the other end of that conduit, wherever it opens up outside the room. Corn: If that exterior opening isn't sealed, you've got a direct path for unfiltered air to travel through the conduit and into the room if the pressure differential ever drops. Herman: The fix is straightforward. On the interior side, where the cable enters the room through the conduit, you pack the opening with firestop putty or duct seal compound. It's a few shekels at any hardware store. The putty molds around the cable, seals the gap, and maintains the pressure boundary. The wall plate goes over it, and you've got a clean Ethernet port with no meaningful leakage. Corn: The advice is: seal the conduit, not the jack. The wall plate is cosmetic. The putty behind it is what's doing the work. Herman: This is where we get to the actual weakest link in a mamad's protection, and it's not the Ethernet jack or the door gasket. It's power. Corn: The overpressure fan runs on electricity. No power, no fan. No fan, no positive pressure. And without positive pressure, the room reverts to its natural state — which is just a leaky concrete box with no protection against airborne contaminants. Herman: Most mamadim have a single electrical outlet. If the grid goes down — and during rocket attacks, power infrastructure is often targeted or simply fails under load — that fan stops. The room doesn't become dangerous instantly, but it loses its CBRN protection the moment the pressure equalizes. Corn: That's a much bigger problem than a six-millimeter hole for an Ethernet cable. The overpressure system is electrically dependent. The filtration unit doesn't have a hand crank. There's no backup bellows. Herman: Compare this to Swiss bunkers, which use a completely different philosophy. Switzerland has enough blast-hardened shelter space for its entire population — it's been national policy since the nineteen sixties. Those bunkers use fully sealed doors with separate, dedicated ventilation systems, often with manual backup pumps. They're designed for nuclear fallout scenarios where you might be underground for days or weeks. Corn: Different threat model, different engineering tradeoffs. The Swiss are preparing for a nuclear exchange that might contaminate the surface for weeks. Israel is dealing with short-range rockets and the possibility of chemical warheads that require hours of protection, not weeks. The mamad doesn't need to be a self-sufficient bunker. It needs to keep you alive until the all-clear. Herman: That distinction only holds if the power stays on. And that's the thing nobody talks about when they're debating wall jacks and door gaskets. The real failure mode isn't a slow leak through a conduit. It's a blackout that kills the fan entirely. Corn: If you're serious about mamad preparedness, the single most impactful thing you can do isn't sealing every microscopic gap. It's putting a UPS — an uninterruptible power supply — on that filtration unit. Herman: A small UPS, maybe a few hundred watt-hours, can keep the fan running for hours after the grid drops. Pair that with a charged phone and a battery-powered radio, and you've addressed the actual vulnerabilities. Not the theoretical ones. Corn: Which brings us back to Daniel and Hannah's Shabbat debate. They were arguing about whether a few millimeters of gap in a wall jack defeats the room's protection. The real answer is that the wall jack is fine, the door gasket is fine, the whole leaky-but-pressurized system is fine — as long as the fan has power. Herman: The Ethernet jack was never the weakest link. The single outlet behind it was. Corn: All right, so we've established the fan is everything and the single outlet is the real vulnerability. Let's get practical. If you're listening to this and you have a mamad, there are maybe three things worth doing this week. Herman: First one is dead simple. Test the overpressure. Close the door, turn on the filtration unit, and hold a tissue near the gap between the door and the frame. Doesn't need to be a lab instrument. Just a tissue. If it flutters outward, the room is pressurized. If it doesn't move, or worse, gets sucked inward, something's wrong — either the filter is clogged or undersized, or the room is too leaky to hold pressure. Corn: If the tissue just hangs there limp, that's the scenario where the fan is running but not overcoming the leakage. You've got no protective gradient. Herman: And that's worth fixing, because without positive pressure you've basically got a very heavy bedroom with a fancy air purifier in the corner. It's cleaning the air that happens to drift through it, but it's not keeping anything out. Corn: Second thing: if you want internet in there, the cleanest approach is power over Ethernet — PoE — from a router that's outside the mamad and on a UPS. One cable carries both data and power. You drill a single penetration through the conduit, pull the cable through, and then seal around it with duct seal putty. It's a few shekels at any hardware store, it never hardens, and it molds around the cable like clay. Herman: The putty goes on the interior side of the penetration, behind the wall plate. That's what maintains the pressure boundary. The wall plate itself is just a plastic cover. And the beauty of PoE is you're not consuming the mamad's single outlet for your access point. The power comes from outside. Corn: Which leaves that precious outlet free for the thing that actually keeps you alive. Herman: Which brings us to the third thing, and this is the one that most people overlook. Get a UPS for the filtration unit. A small one, even a basic model with a few hundred watt-hours, can keep the fan running for hours after the grid drops. The fan doesn't draw much power — it's a blower motor, not a space heater. Corn: If you're thinking about extended outages — and in certain scenarios that's not hypothetical — you can step up to a twelve-volt deep-cycle battery with an inverter. That'll run the fan and charge devices for a full day or more. Herman: The filtration unit is the single point of failure for the entire protection concept. Everything else — the door gasket, the wall jacks, the window seal — all of it depends on that fan spinning. A UPS turns a vulnerability into a solved problem for about the cost of a nice dinner. Corn: This is where the broader takeaway lands, I think. Daniel and Hannah were debating whether the room could be truly protective given visible imperfections. But the engineering answer is that perfect is the enemy of good. A mamad that keeps you alive for four hours with internet access and filtered air is more useful than a theoretically hermetic bunker you can't stand to be in. Herman: The Swiss bunker approach — perfect seals, airlocks, manual backup ventilation — that's a marvel of engineering for a weeks-long nuclear fallout scenario. But it's also a miserable place to spend an afternoon with a five-year-old. The mamad makes a different trade. It sacrifices theoretical perfection for practical habitability. Corn: That trade is visible everywhere once you know to look for it. The gaps that let air cycle. The single outlet that forces you to think about power. The wall jacks that let you stay connected to the outside world. None of it is an accident. It's all designed around a specific threat model and a specific use case. Herman: Which is short-range rockets and the possibility of chemical or radiological contamination, with a stay measured in hours not days. Within that envelope, the mamad is genuinely clever engineering. Outside that envelope — extended power outage, nerve agents in vapor form, a nuclear blast close enough to compromise the structure — the protection degrades or fails entirely. Corn: Knowing where those boundaries are is more valuable than either assuming the room is invincible or assuming the gaps make it useless. Both of those are wrong in ways that could actually matter. Corn: One question that sticks with me after all of this — and it's bigger than the mamad itself — is whether overpressure-filtered rooms start showing up in places that have nothing to do with rockets. Think about what's happening with climate. Wildfire smoke events that turn the sky orange for days. Dust storms that make breathing outdoors dangerous. We saw it in New York last summer, we've seen it in Sydney, in Beijing, in Athens. Herman: The same principle applies. A room running positive pressure with HEPA filtration keeps smoke particulates out the same way it keeps fallout dust out. The threat is different but the physics is identical. And I could absolutely see building codes in fire-prone regions starting to require exactly this — not a blast-hardened concrete box, but a designated room with a filtered positive-pressure system. Corn: It's already happening in pieces. California has "clean air shelters" in some communities. Australia experimented with them during the Black Summer fires. But they're public facilities, not residential standards. The mamad model — every apartment has one, it's just part of the floor plan — that's a different level of normalization. Herman: The cost isn't astronomical once you're already building. The filtration unit, the slightly better door gasket, the pressure-rated window. It's incremental on new construction. The expensive part is retrofitting, which is why you'd want it in the building code from the start. Corn: The Israeli safe room might end up being an accidental export — not because anyone else is worried about rockets, but because the engineering solution to one problem turns out to solve a completely different one that half the world is about to have. Herman: Which is how good engineering often works. You design for a specific threat, and the principle turns out to be portable. Corn: That brings me back to Daniel and Hannah's Shabbat debate. They were arguing about Ethernet jacks and door seals, but the real question underneath it was whether you can trust a system that looks imperfect. Whether gaps you can see mean the protection isn't real. Herman: The answer turns out to be yes — if you understand the principle behind the design. The gaps aren't evidence of failure. They're evidence of a different strategy. The room doesn't keep bad air out by being a fortress. It keeps bad air out by being a fountain. Clean air flows out, and nothing flows back in. Corn: Once you see it that way, the visible imperfections stop being alarming and start being legible. The gap around the door isn't a flaw. It's the exhaust port. The wall jack isn't a breach. It's just another place where the air says "out, not in. Herman: That's a pretty good metaphor for a lot of things, actually. The systems that look the most bulletproof on the surface are often the ones that fail catastrophically when something unexpected happens. The systems with visible gaps and compensating mechanisms — those are the ones that keep working. Corn: Daniel and Hannah were both right about what they saw. They just didn't have the one piece of information — overpressure — that made their observations fit together instead of contradict each other. And that's the whole show, really. Two people looking at the same room, seeing different things, both correct, and the truth is the thing that connects them. Herman: If you enjoyed this episode, please leave a review — it helps other curious people find the show. And send us your weird prompts at my weird prompts dot com. Corn: Now: Hilbert's daily fun fact. Hilbert: In the early medieval period, scholars in the Malian city of Timbuktu attributed the pink color of a local salt lake to a species of bacterium that supposedly fed on gold deposits — a claim later corrected when researchers identified it as a halophilic archaeon producing carotenoid pigments to protect itself from intense UV radiation. The gold-feeding part was entirely made up by a fourteenth-century traveler trying to impress a patron. Corn: A medieval influencer invented extremophile microbiology for clout. Herman: Some things never change. This has been My Weird Prompts. We'll catch you next time.