Latest / Logosofia no Dia a Dia / Tall vs Short: Which Floor Survives an Earthquake?
Transcript
- [Prompt] custom: I love to learn about what's involved in earthquake proofing buildings. Or even earthquake mitigatin | Hosts: corn, herman [Response] Corn: Daniel sent us this one — he's been thinking about earthquake-proofing, and honestly, the whole thing starts with a question that sounds like it should have an obvious answer but doesn't. If a magnitude seven point zero quake hit Tel Aviv tonight, are you safer on the second floor or the thirtieth floor? Herman: Most people would say the second floor, and most people would be wrong. Or at least, it depends on about six different things we're going to get into. Corn: The reason this isn't just an academic question — especially for Daniel, for anyone listening in Israel — is that the Dead Sea fault is statistically overdue for a major event. The last big one was nineteen twenty-seven, magnitude six point two, killed about five hundred people. We're coming up on a hundred years. Herman: And Israel updated its seismic code, Standard four thirteen, in twenty twenty-four. So we now have the most current engineering requirements in place, but the question is whether the buildings people actually live in were built to them. Corn: Most weren't. We'll get to that. But the core of what Daniel's asking is really two things. One, is earthquake-proofing even a real thing, or is it marketing? And two, what do those fortifications in Israeli buildings actually involve? Because you walk past construction sites, you see the rebar, you hear the term "shelter room" thrown around, but the seismic engineering underneath all that is a whole different layer. Herman: The short answer to "is earthquake-proofing real" is — no. Not in the way most people mean it. Engineers don't design buildings to be undamaged after a major quake. They design them so you walk out alive. The technical term is "life safety." The building might be a total loss. It might be leaning at a fifteen-degree angle. But it doesn't pancake, and you survive. Corn: Which is a very different promise than the word "proof" implies. Herman: Earthquake-proof is a real estate term, not an engineering one. Corn: We're going to spend this episode unpacking two big misconceptions. The first is that shorter buildings are always safer — and the physics of that is genuinely surprising. The second is that earthquake-proofing is one technology, one thing you bolt on, when actually it's a layered system of design choices that go all the way down to the foundation. Herman: For Israel specifically, we'll walk through what Standard four thirteen actually requires, what base isolation is and why it's barely used here, and what you should know if you're living in an apartment building that was built before anyone was thinking about seismic loads. Corn: Which, spoiler, is most of them. Herman: Let's sit with that "life safety" idea for a second, because it reframes everything. When an engineer says a building meets code, they're not saying it'll be habitable after a magnitude seven quake. They're saying the probability of anyone dying inside is below a certain threshold. That threshold, in Israel, is baked into Standard four thirteen — the building must survive a quake with a four hundred seventy-five year return period without collapsing. That's the ten percent probability of exceedance in fifty years benchmark. Corn: Which sounds precise, but also — a four hundred seventy-five year quake could happen next Tuesday. The return period is a statistical average, not a schedule. Herman: And that's what keeps structural engineers up at night. You're designing for probabilities, but earthquakes don't care about probabilities. So the entire discipline is built around layers of redundancy. If one system fails, the next one catches it. If the columns crack, the shear walls take over. If the shear walls fail, the ductile frame redistributes the load. No single point of failure. Corn: That's the second misconception Daniel raised, even if he didn't phrase it exactly this way. People picture earthquake-proofing as a thing — base isolators, or some kind of shock absorber you install. But it's not a product. It's a philosophy of design that touches every decision from the soil report to the roof. Herman: The philosophy has a name. It's called capacity design. The idea is you deliberately choose which parts of the building will fail first, and you make sure those parts fail in a ductile way — bending, stretching, absorbing energy — rather than a brittle way, which is snapping suddenly. You want the building to groan, not shatter. Corn: Which is a weird mental model to sit with. You're engineering a controlled failure. Herman: It's like a pediatrician managing a fever. You don't prevent the fever, you manage how the body responds so the outcome is recovery, not seizure. The building goes through the earthquake, it deforms, it dissipates energy through those designated plastic hinge zones, and at the end it's standing. Ugly, maybe condemned, but standing. Corn: When we talk about "earthquake-proofing" from here on, what we actually mean is seismic design for life safety. And the reason the tall-versus-short question is so counterintuitive is that it depends entirely on what kind of ground the building sits on and what frequency the earthquake is shaking at. Herman: There's a third misconception lurking underneath the two Daniel named, which is that building codes are static. They're not. Standard four thirteen was first introduced in nineteen seventy-five, revised significantly in nineteen ninety-five, and updated again just two years ago. Each revision reflects lessons learned from somewhere else's disaster. The nineteen ninety-five update came after the Northridge quake exposed weaknesses in welded steel moment frames that everyone had assumed were fine. The twenty twenty-four update added requirements for site-specific response spectra — meaning you can't just use a generic earthquake model anymore, you have to analyze the actual soil under your specific building site. Because soil isn't just dirt. It's an amplifier. Or a dampener. Depends on the frequency. Corn: Let's talk resonance. Every building has a natural frequency — the rate at which it wants to sway back and forth. For a thirty-story tower, it's typically between two and five seconds per cycle. For a three-story walk-up, maybe point two to point five seconds. Very fast, very stiff. Herman: Now here's where it gets interesting. Earthquakes shake the ground at a range of frequencies, but the soil underneath filters and amplifies certain frequencies over others. Deep, soft soil — like the Tel Aviv coastal plain, layers of sand and clay going down tens of meters — that soil has its own natural period, typically around one to two seconds. Corn: You've got the building's frequency and the soil's frequency, and if those two match the earthquake's dominant frequency, you get resonance. Which is what, exactly? Herman: Resonance is when you push a swing at exactly the right moment and the arc gets bigger and bigger with each push. In a building, each seismic wave that arrives at the resonant frequency adds energy instead of dissipating it. What might have been a gentle nudge becomes a violent lurch. This isn't theoretical — it's exactly what happened in Mexico City in nineteen eighty-five. The quake was four hundred kilometers away, off the Pacific coast. But Mexico City is built on an ancient lakebed — incredibly soft, water-saturated clay. That clay amplified the long-period waves to about a two-second period. And the buildings that collapsed disproportionately? Six to fifteen stories. Buildings shorter than six mostly survived. Taller than fifteen mostly survived. The mid-rise buildings — the ones whose natural period matched that two-second soil resonance — they came down. Corn: The tall buildings were actually safer. Herman: In that specific context, yes. Their natural period was four, five seconds — far enough from the soil's resonant frequency that they didn't amplify. The short buildings were too stiff to resonate at all. It was the middle band that got destroyed. Corn: Which completely upends the intuition that closer to the ground is always better. Herman: It gets weirder. Flip the scenario. If the epicenter is close — say, a quake on the Dead Sea fault right under Jerusalem — the dominant shaking is high-frequency, short-period jolts. In that case, short stiff buildings resonate with the ground motion and tall flexible buildings don't. So the "which is safer" question has no single answer. It depends on distance from the fault, soil type, and the building's own period. It's a matching problem, not a height problem. Corn: There's the whiplash effect. Short buildings on soft soil can experience amplified acceleration because the soil's resonant frequency matches the building's own short period. Herman: Loma Prieta, nineteen eighty-nine. Magnitude six point nine. The Marina district in San Francisco is built on fill — loose, sandy soil that used to be a lagoon. When the shaking started, that soil amplified the motion, and in some places it liquefied. Two-story wood-frame apartment buildings pancaked. Meanwhile, a twenty-story reinforced concrete tower a few blocks away was essentially undamaged. The tower's long period didn't match the soil's short-period amplification, and it was engineered with ductility in mind. Corn: Which brings us to stiffness versus ductility. A building that's too rigid will resist deformation right up until it can't, then fail suddenly. Think of a piece of chalk. Herman: A ductile building bends. Steel reinforcement stretches. Concrete cracks, but the cracks are controlled — hairline, distributed — not catastrophic. The building absorbs enormous amounts of energy through that deformation. It comes out permanently tilted, maybe, with doors that don't close, but it doesn't collapse. Corn: The goal isn't to resist the earthquake. It's to ride it out by yielding in controlled ways. Herman: That's capacity design. You designate specific zones — plastic hinges — usually at beam ends, sometimes at the base of shear walls. You reinforce those zones heavily with closely spaced steel stirrups, so when they yield, they do it gradually. Meanwhile, the columns are designed to be stronger than the beams. Strong column, weak beam. So the hinges form in the beams, not the columns. Corn: Because if a column fails, that floor comes down. Herman: If a column fails, the entire load path is compromised. One column buckles, the floor above drops, the impact overloads the columns on the next floor down, and it cascades. The strong-column-weak-beam principle prevents that chain reaction. The beams can fail, the building can lean, but the vertical structure stays intact. Corn: All of this is invisible. You walk into a building, you see drywall and tile. You have no idea whether the rebar inside the columns was spaced at ten centimeters or twenty. Whether the stirrups hook at a hundred thirty-five degrees or ninety. Those details are the difference between walking out and being pulled from the rubble. Herman: The hundred thirty-five degree hook is a perfect example. In a seismic event, the concrete cover spalls off. If the stirrup ends are bent at ninety degrees, they pop open and the column core crumbles. A hundred thirty-five degree hook locks into the core concrete. It's a tiny geometric detail that costs almost nothing during construction, and it's the difference between a column that holds and a column that disintegrates. Corn: When we talk about earthquake-proofing, what we're really talking about is hundreds of those tiny geometric details, layered together, any one of which can be the thing that fails. Herman: The soil underneath all of it is the first domino. If you don't understand what the ground is doing, none of the structural design matters. Corn: Let's get concrete about what Israel actually requires. Standard four thirteen — what does it actually mandate? Herman: Three main strategies that work together. First, ductile reinforced concrete frames. The steel rebar inside the concrete is detailed with closely spaced stirrups, those hundred thirty-five degree hooks, all concentrated at beam-column joints where stress is highest. In a critical zone near a joint, you might have stirrups every ten centimeters. Outside that zone, maybe every twenty. If a contractor spaces them all at twenty because it's easier, the column looks identical once the concrete is poured, but it will fail in a quake. Corn: There's no way to inspect that after the fact without X-raying the concrete, which nobody does. Herman: Which is why enforcement during construction is everything. Second strategy: shear walls. Continuous reinforced concrete walls that run from foundation to roof, usually around elevator shafts or stairwells. They act as the building's spine, taking lateral forces and transferring them down to the foundation. Third: strong column, weak beam. A deliberate hierarchy of failure. Corn: All of this is mandatory now? Herman: For any building permitted after the code was adopted, yes. But here's the timeline that matters. Pre-nineteen seventy-five — no seismic provisions at all. The building was designed for gravity loads only. Seventy-five to ninety-five — basic seismic provisions, but relatively crude by modern standards. Post-ninety-five — the code got serious, incorporating lessons from Northridge and Kobe. And the twenty twenty-four update added site-specific soil analysis. If you're building on the coastal plain, the soft soil response spectrum is different from building on Jerusalem limestone. Corn: If you're in a Tel Aviv apartment tower built in nineteen ninety-eight, you're in decent shape. If you're in a Jerusalem walk-up from nineteen sixty-two, not so much. Herman: The nineteen sixty-two building has zero seismic design. And Jerusalem is closer to the fault. Corn: Which brings us to base isolation. Everything we've described so far — ductile frames, shear walls, strong columns — those are about making the building survive the shaking. Base isolation is about making the building not shake in the first place. Herman: You essentially put the building on giant flexible bearings — layers of rubber bonded to steel plates — between the foundation and the structure above. When the ground lurches sideways, the bearings deform and absorb that motion. The building above just sits there. And critically, the bearings shift the building's natural period way out — three to five seconds — far beyond where most earthquake energy is concentrated. Corn: This is common in Japan. They require it for all buildings over sixty meters. Herman: In Israel, it's still rare. The new Tel Aviv light rail control center uses it, because that's a critical facility that needs to function immediately after a quake. But for residential towers? It adds maybe five to ten percent to construction cost, and developers here have been reluctant. Corn: Is that a regulatory failure or an economic one? Herman: It's a risk tolerance question. Japan sits on the Ring of Fire. They get magnitude seven quakes every few years. Israel's major quakes are separated by decades or centuries. The calculus is different. But the twenty twenty-three Turkey-Syria earthquakes made the case better than any engineering paper could. Buildings that met modern codes survived and stayed operational. Pre-code buildings pancaked. Over fifty thousand dead, and the single biggest predictor of survival wasn't height or location. It was code compliance. Corn: The code works. When it's actually followed. Herman: When it's enforced. Turkey had a modern seismic code on the books before those quakes. It just wasn't enforced. Builders cut corners, inspectors looked the other way, and the buildings that collapsed were full of people who assumed someone had done their job. Corn: Which is the question hanging over every conversation about this in Israel too. Having Standard four thirteen is one thing. Whether the rebar spacing in your building's columns actually matches the drawings is another. Herman: Let's translate all of this into something practical. If you're listening in Israel and trying to figure out what this means for the building you actually live in, there's a timeline that matters more than almost anything else. Corn: The construction year. Herman: Pre-nineteen seventy-five — designed for gravity only. Those are the buildings that make engineers nervous. Seventy-five to ninety-five — basic seismic provisions, but early generation. Stirrup spacing requirements were looser, soil amplification understanding was primitive, and strong-column-weak-beam wasn't consistently applied. Post-ninety-five, you're in pretty good shape. The code got serious. And the twenty twenty-four update is best practice. Corn: If you're renting or buying, that's the first thing to check. Not the height, not the view. What year was the concrete poured? Herman: If it's pre-seventy-five, that doesn't mean you should panic and move tomorrow. But it does mean you should know. And if you're on the building committee, if you're an owner — there are options. Corn: Which brings us to the second practical thing. In an actual earthquake, what do you do? Because the instinct is to run outside. Herman: That instinct kills people. The greatest danger in a modern building isn't structural collapse — it's falling debris. Facade panels, parapets, broken glass. The safest place is under a sturdy table, away from windows, holding on. Drop, cover, hold on. It's not glamorous advice, but it's what the data supports, quake after quake. Running outside puts you directly in the path of everything coming off the building. And in a pre-code building, you still might not have time to get out before the shaking intensifies. The P-waves arrive first as a warning. The destructive S-waves come seconds later. By the time you feel the strong shaking, you're better off where you are. Corn: The advice is boring but it saves lives. Drop, cover, hold on. Herman: Third practical thing, for homeowners specifically. If you own an apartment in an older building, you're not helpless. Three main approaches: adding shear walls to stiffen the structure, wrapping existing columns with carbon fiber or steel jackets to improve ductility, and installing steel bracing frames on the exterior. Corn: Israel has a program for this. Herman: TAMA thirty-eight. It's an urban renewal framework that gives developers incentives to strengthen older buildings. A developer reinforces your building, adds a safe room to every apartment, and in exchange gets the rights to add floors on top. The seismic retrofit gets paid for by the new construction. Corn: Which sounds almost too good to be true. Free earthquake reinforcement in exchange for letting someone build a penthouse? Herman: It's not free — the developer profits from the new units. But for existing residents, the out-of-pocket cost can be zero or very low. The catch is that TAMA thirty-eight only applies to buildings permitted before nineteen eighty, and the process requires agreement from a supermajority of residents. It can take years. Corn: In the meantime, the building sits there unreinforced. Herman: Which is why the government has been trying to streamline the approval process. There's also TAMA thirty-eight slash two, which allows for full demolition and rebuild instead of just retrofitting. But that's a much bigger undertaking. Corn: The bottom line: know your building's year, know what to do when the shaking starts, and if you're in a pre-eighty building, find out whether TAMA thirty-eight is on the table. Herman: There's a question that sits underneath all of this, though, and I think it's where things get interesting over the next decade. We've been talking about designing buildings to survive earthquakes. But what if the building knew the earthquake was coming before you did? Corn: That's already happening. Japan's Shinkansen trains have had seismic detection for years. Sensors along the coast pick up P-waves and the system cuts power to the trains before the damaging S-waves hit. Ten to thirty seconds of warning. Herman: For a building, same principle. You install accelerometers at the foundation. When they detect P-waves, elevators automatically stop at the nearest floor and open doors. Gas lines shut off. Backup generators spin up. In more advanced setups, active mass dampers — essentially giant pendulums near the top of the building — can shift position to counteract the expected sway. The building braces itself. Corn: The question is whether any of that makes it to Israeli buildings. Herman: The technology isn't the barrier. The sensors exist, the algorithms exist. Japan, Taiwan, parts of California already have building-level early warning integration. For Israel, the challenge is more about whether the investment makes sense given the recurrence interval. A thirty-second warning system is life-saving when quakes happen every few years. When they're separated by a century, it's harder to justify the maintenance cost of keeping those sensors calibrated and networked. Corn: Though the cost curve on sensors has basically collapsed. A decent triaxial accelerometer costs maybe a few hundred dollars now. Herman: The expensive part isn't the hardware, it's integrating it with building control systems and maintaining the network. But I could see it becoming standard in new high-rises within ten years, especially if the twenty twenty-four code update opens the door for performance-based design incentives. Corn: Which is the other shift you mentioned. Performance-based design. That's the move from "don't collapse" to something more nuanced. Herman: Traditional code is prescriptive. It says: use this much rebar, space stirrups this far apart, make columns this strong. Follow the recipe. Performance-based design flips that. You define the outcome you want — immediate occupancy, life safety, or collapse prevention — and the engineer designs to meet that performance target for a specific earthquake level. A hospital owner can say, "I want this fully operational after a two-thousand-year quake." A warehouse owner can say, "I just need it not to kill anyone. Corn: You're buying a specified level of resilience rather than just checking code boxes. Herman: It changes the economics. Under prescriptive code, every building gets roughly the same level of protection regardless of what's inside. Performance-based design lets you allocate resources where they matter most. Critical infrastructure gets higher performance targets. Low-occupancy buildings get the minimum. It also makes retrofitting more feasible — if you can't afford to bring a nineteen-sixties apartment block to full modern code, maybe you can bring it to a defined partial-performance level that keeps people alive. Corn: That's the arc of seismic engineering in one sentence. From "earthquake-proof is a lie" to "you can actually choose your level of unproofedness with surprising precision. Herman: The thing I keep coming back to is that none of this is secret knowledge. The physics is well understood. The detailing requirements are published. The case studies — Mexico City, Loma Prieta, Turkey — they all point to the same conclusions. The gap isn't between what we know and what we need to discover. It's between what's in the code and what's in the ground. Corn: Which is a good place to leave it. Daniel, hopefully that answers the question — and hopefully your building was built after ninety-five. Herman: Now: Hilbert's daily fun fact. Hilbert: The Tuyuca language of the Amazon requires speakers to grammatically mark how they know every piece of information they report — whether they saw it, heard it, inferred it, or were told by someone else. Linguists call this evidentiality marking, and an unintended consequence is that Tuyuca speakers are extremely difficult to lie to in their native language, because the grammar itself demands you reveal your source. Corn: You can't even say "the fish was this big" without the verb ratting you out. Herman: A language with built-in citation requirements. Hilbert, that's unsettling. Corn: Thanks to Hilbert Flumingtop for producing, and thanks to everyone listening. If you found this useful, leave us a review wherever you get your podcasts — it actually helps. This has been My Weird Prompts. We'll be back next week.