Daniel sent us this one — he's been watching Jerusalem get dug up, seeing all the pipes and cables beneath the pavement, and it got him thinking. He points out that cities have enormous clustered demand for electricity, yet we don't see power pylons marching through downtown. So where is it all? He assumes underground. But then he asks two sharp questions. First, when they're laying high-voltage cable today, how do they know with centimetric precision where to put it? And second, how much of this transition is actually happening — are all those ugly overhead wires for phone and power being migrated underground so we get clean streetscapes? There's a safety angle, an engineering angle, and a city-planning angle. It's a good one.
It's a fantastic question, and the short answer is that in most modern cities, the high-voltage transmission lines — the big pylons — do stop at the city boundary. They feed into substations, and from there, distribution happens through a network that's overwhelmingly underground in dense urban cores. But Daniel's right to notice it's not universal. You still see overhead wires in plenty of places, and the reasons for that are a mix of cost, geology, and legacy infrastructure that's incredibly expensive to move.
By the way, quick note — DeepSeek V four Pro is writing our script today, so if anything sounds unusually articulate, that's why.
I'll take that as a compliment to our usual standards. So let me start with the precision question, because it's the one that has the most interesting engineering behind it. Daniel asked about centimetric precision for placing high-voltage cables. The reality is we now have multiple overlapping systems that locate underground utilities with accuracy down to a few centimeters. Ground-penetrating radar is the workhorse — it sends electromagnetic pulses into the ground and reads the reflections off different materials. units can detect cables, pipes, and voids at depths down to about ten meters, with horizontal resolution in the centimeter range.
It's sonar for dirt, basically.
More like radar for dirt, but the principle's similar. The key thing is that G. by itself isn't enough for what Daniel's describing. When you're actually installing new high-voltage cable in a city street, you're not just locating things — you're coordinating with a dozen different utilities, each of which has its own records of what's down there. And those records vary wildly in quality depending on when the infrastructure was laid.
That's the part that worries me. You're digging into a Jerusalem street — a city that's been continuously occupied for thousands of years — and you're trusting someone's records from maybe the nineteen fifties?
That's exactly the problem, and it's why the industry has moved toward something called S. — Subsurface Utility Engineering. It's a formal process with four quality levels, from A to D. Level D is just existing records, which as you say are often wrong. Level C is visual site inspection. Level B uses geophysical methods like G. and electromagnetic locators. And Level A is actual physical exposure — you dig test holes, which they call potholing, using vacuum excavation.
That sounds dramatic.
It's actually quite gentle. Instead of a backhoe that could slice through a cable, they use high-pressure air or water to break up the soil and a vacuum truck to suck it out. It exposes the utility without touching it. When you're dealing with a hundred and sixty-one kilovolt transmission cable buried under a city street, you really, really want to know exactly where it is before your digger bucket arrives.
The precision comes from this layered approach — records plus geophysics plus physical verification — not from one magic device that sees everything.
And there's another layer that's becoming standard for new installations. They're embedding tracer wires and radio-frequency identification tags directly into cable conduits now. The tracer wire carries a signal that a surface receiver can detect, giving you the exact path of the cable. tags store information about the cable's specifications, installation date, depth, voltage — and they're readable from the surface with the right equipment. So fifty years from now, someone digging in that spot won't have to rely on paper records from twenty twenty-six.
That assumes the R. tags still work in fifty years.
tags don't have batteries. They're energized by the reader's signal. As long as the metal doesn't corrode, they'll function essentially forever. The bigger risk is that nobody in twenty seventy-six knows to look for them. But that's a standards and documentation problem, not a technology problem.
Let me pull us back to Daniel's other question, because I think it connects. He's looking at Jerusalem's overhead wires — phone lines, maybe some power distribution — and asking whether all of this is being migrated underground. Is that the direction we're heading globally?
The direction, yes. The speed, wildly uneven. Let me give you some numbers. In much of Western Europe — Germany, the Netherlands, Denmark — more than eighty percent of urban distribution lines are already underground. In parts of the Netherlands, it's above ninety-five percent. New developments there haven't even considered overhead lines for decades.
On the other end?
In the United States, it's roughly thirty-five to forty percent underground in urban areas, but that varies enormously by region. Older northeastern cities like New York and Boston have extensive underground networks because they were built out in the early twentieth century when undergrounding was seen as a civic necessity. But huge swaths of suburban and even urban America still run on overhead lines, and the cost to bury them is staggering.
How staggering are we talking?
Per mile, underground residential distribution costs somewhere between five hundred thousand and two million dollars, compared to about a hundred and fifty to three hundred thousand for overhead. In dense urban areas with existing underground congestion, it can go much higher. I saw a report from a California utility estimating that a comprehensive undergrounding program for their service area would cost over forty billion dollars and take decades.
A city like Jerusalem, which has ancient archaeology layered under modern infrastructure, would be on the extreme high end of that cost curve.
Every shovel of dirt in Jerusalem potentially hits something archaeologically significant, which triggers a legal requirement to stop work and bring in the Israel Antiquities Authority. That can delay a project by months or years. It's one reason Jerusalem's infrastructure modernization has been slower than, say, Tel Aviv's.
Daniel mentioned seeing the city being dug up right now. So something is happening.
It is, and this is actually a significant moment for Jerusalem's infrastructure. The city has been undergoing a major overhaul of its underground utilities — water, sewage, gas, and power — coordinated through a municipal master plan that tries to do everything at once so they're not digging up the same street three times. The power component involves burying medium-voltage distribution lines that previously ran overhead in some neighborhoods, and upgrading the underground transmission corridors that feed the city center.
That coordination point is worth underlining. The nightmare scenario for a city is what they call "pavement cuts" — each utility digging independently, patching the road, and the next utility digging it up six months later. Good municipal management requires what's sometimes called a "common trench" policy.
And common trench policies are where you really see the centimetric precision Daniel asked about in action. You're designing a trench that might be two meters wide and three meters deep, and you're stacking utilities vertically with precise separation distances. Power cables go deepest, typically at about one point two to one point five meters. Telecommunications above that, at maybe sixty to ninety centimeters. Gas lines at their own specified depth with horizontal separation from electrical. Water mains somewhere in the mix. Every one of those placements is specified by code, and the separation distances are critical for safety.
The safety angle is what I want to push on. Daniel explicitly mentioned concern for safety with high-voltage underground cables. What's the actual risk profile compared to overhead?
It's a trade-off. Overhead lines have higher risk of faults from weather — wind, ice, falling trees, vehicle collisions with poles. Underground cables are protected from all of that, but when they do fail, the failure can be more catastrophic and harder to locate. A cable fault underground can create a voltage gradient in the surrounding soil — step potential, they call it — that can be lethal to anyone standing nearby.
That's the thing where the electricity spreads through the ground and your feet are at different voltages?
Yes, and the current flows up one leg and down the other. It's a real hazard, which is why underground cable installations include extensive grounding and bonding, and why the precise placement matters so much. You need to know exactly where that cable is so that nobody accidentally digs into it, and so that the surrounding infrastructure is properly insulated from it. Modern high-voltage underground cables typically use cross-linked polyethylene insulation — X. — which is remarkably durable. But nothing is indestructible.
The safety concern drives the precision requirement. It's not just about not hitting it during construction — it's about knowing where it is for the entire lifecycle.
That lifecycle is getting longer. Modern underground cables are designed for forty to fifty years of service, but many of the cables being installed now use materials and construction techniques that could push that to sixty or seventy. So the burial depth, the conduit material, the warning tape placed above the cable — all of that has to be documented and locatable decades from now.
Let's talk about the conduit. When Daniel sees a street cut open and there's a bundle of pipes, what's actually in there?
For power, it's typically a high-density polyethylene conduit — H. — bright orange in most countries as a visual warning. The cable itself isn't buried directly in the ground in modern installations. It's pulled through the conduit, which protects it from soil chemistry, rocks, and future excavation. The conduit also allows the cable to be replaced without digging up the street again — you pull the old cable out, pull the new one through. That's a huge advantage for maintenance.
The orange color is a standard?
In most of the world, orange is power. The American Public Works Association — A. — standardized a color code that's been widely adopted internationally. Red is electric power lines, actually — I should correct that. Orange is communications. Red is electric. Yellow is gas. Blue is water. Green is sewer. Purple is reclaimed water. White is proposed excavation. So when you see a street marked up with spray paint in different colors, that's what's happening — they're marking the horizontal path of every utility below.
That's the kind of standardization that nobody appreciates until it fails.
It fails more often than you'd hope. The Common Ground Alliance in the U. tracks utility damage incidents, and their most recent data shows something like four hundred thousand excavation-related utility damages per year in the United States alone. That's more than a thousand per day. The leading cause is failure to notify the one-call center before digging. The second leading cause is inaccurate facility marking.
The precision tools exist, but the human and organizational layer is where it breaks down.
That's the story of infrastructure in general. The technology is rarely the limiting factor. It's coordination, record-keeping, and the willingness to pay for thoroughness.
Let me circle back to Daniel's question about clean streetscapes. He mentioned those ugly overhead wires — he guessed they might be legacy phone connections. Is the aesthetic argument actually driving undergrounding?
It's a surprisingly powerful driver, and it's not just aesthetics. There's a concept in urban planning called "visual pollution" or "streetscape clutter," and overhead utilities are one of the biggest contributors. Cities that have invested heavily in undergrounding — like London's Kensington and Chelsea, or much of central Paris — have seen measurable increases in property values. There's a study from the University of Wisconsin that found home values increased by about three to five percent on streets where overhead lines were buried.
Three to five percent is real money.
It is, and when you multiply that across a neighborhood, the aggregate property value increase can sometimes justify the cost of the undergrounding project on its own, at least in high-value urban areas. But in lower-density or lower-income areas, the math doesn't work, which creates an equity problem. Rich neighborhoods get clean streetscapes and higher property values. Poor neighborhoods keep the wires.
That's the kind of thing that should make municipal planners uncomfortable.
It does, and it's sparked debates about whether undergrounding should be treated as a basic infrastructure investment rather than an aesthetic upgrade. If you frame it as reliability — underground lines are protected from storms — then it becomes a resilience investment, which is easier to justify with public funds.
How much more reliable are we talking?
Overhead lines experience about nine times more outage minutes per customer per year than underground lines, according to data from the Edison Electric Institute. Most of that difference is weather-related. In hurricane-prone regions, the gap is even larger. After Hurricane Sandy in twenty twelve, there was a huge push in the northeastern U. to underground critical distribution lines. Some of that happened, but the cost estimates killed a lot of the ambitious proposals.
Nine times more outages. That's a number that should be in every cost-benefit analysis.
It should, but you have to factor in that when underground lines do fail, the repair time is much longer. An overhead line fault can often be fixed in hours. An underground fault can take days to locate and repair, because you have to find the fault location — often with specialized equipment that analyzes the reflection pattern of a pulse sent down the cable — and then you have to dig it up.
The reliability argument is about frequency of outages, not duration when they happen.
And that's the trade-off utilities are constantly weighing.
Daniel also mentioned something interesting — he said he presumes the power is underground, but he's not sure. That uncertainty is worth sitting with for a moment, because I think a lot of people share it. We walk through cities, we see overhead wires, we don't really know what's in them.
The answer varies enormously by location. In Jerusalem, the high-voltage transmission — the stuff coming from the national grid at a hundred and sixty-one kilovolts — is mostly underground within the city limits. The Israel Electric Corporation has been systematically undergrounding its urban transmission for decades. But the lower-voltage distribution — the lines that actually connect to buildings — is a mix. Some neighborhoods are fully underground. Others still have overhead distribution on poles, often sharing those poles with telecom lines.
The shared poles are the messy ones, visually.
They are, and they're messy from a management perspective too. In many cities, the poles are owned by the electric utility, and telecom companies lease space on them. That creates a coordination nightmare when you want to underground the power lines — you can't just cut the telecom lines, so you have to get every company that has equipment on those poles to simultaneously transition to underground. It's like herding cats, except the cats are regulated monopolies with competing interests.
Is there a city that's done this particularly well as a model?
Tokyo is often cited as the gold standard. After the nineteen ninety-five Kobe earthquake, which brought down thousands of overhead lines and hampered emergency response, Japan launched an aggressive undergrounding program. Tokyo now has over ninety percent of its urban distribution underground. But Tokyo also has the advantage of being rebuilt substantially after World War Two, and the density makes the per-customer cost much more manageable.
Density helps the economics?
The cost per mile of undergrounding is high, but the cost per customer depends on how many customers are served per mile. In a dense urban neighborhood with apartment buildings, you might have thousands of customers per mile of cable. In a suburban area with single-family homes, you might have fifty. The per-customer cost in the suburb is twenty to forty times higher. That's why you see undergrounding concentrated in city centers and largely absent in rural areas.
Daniel's experience in Jerusalem — a dense, historically layered city — is almost a perfect case study in the tensions we're describing. High density makes undergrounding economically more viable. Ancient archaeology makes it logistically much harder. And the visual impact of overhead wires in a city of stone and history is particularly jarring.
Jerusalem is also a city where the aesthetic argument has unusual force. It's a UNESCO World Heritage site in its old city. The visual character of the streetscape isn't just a matter of property values — it's tied to tourism, cultural heritage, and national identity. There's been a concerted effort to underground utilities in and around the Old City specifically to preserve the visual integrity.
Which brings us to the transition question Daniel asked. Is everything being migrated underground? The honest answer is no — not everything, not everywhere, not quickly. But the trajectory is clear. New construction is almost universally underground in developed countries. Major urban renewal projects include undergrounding as a standard component. And the technology for locating and managing underground infrastructure keeps improving.
There's a technology I haven't mentioned yet that's becoming increasingly important: geographic information systems — G. — that create three-dimensional digital maps of underground infrastructure. Cities like Singapore and Helsinki have invested heavily in creating unified underground asset maps that combine data from every utility, every archaeological survey, every geological borehole. When a contractor wants to dig, they can query the system and get a three-dimensional view of what's below, with accuracy that's constantly being refined.
Singapore I understand — they're famously data-driven and have limited land.
Helsinki has something called the Underground Master Plan, which is genuinely visionary. They've mapped the entire subsurface of the city and designated specific depth zones for different uses — transportation tunnels at one depth, utilities at another, geothermal wells at another. They're essentially treating underground space as a finite resource to be planned and allocated, like land on the surface. It's the kind of long-term thinking that most cities only adopt after a crisis.
Most cities are still in the "dig and hope" phase of subsurface management.
The technical term is "chaotic excavation." And it's not just a developing-world problem. London has famously poor subsurface records in some areas, partly because so much infrastructure was laid before standardized record-keeping existed. There are still "ghost" utilities under London — pipes and cables that appear on no map, were abandoned decades or centuries ago, and are discovered only when someone digs into them.
That's a fantastic term. I assume they're not literally haunted.
They're haunted by the ghost of infrastructure past. And they can be dangerous. An abandoned gas pipe that was never purged, an old electrical conduit that was never de-energized — these things lurk under city streets and occasionally make themselves known in dramatic fashion.
When Daniel sees Jerusalem being dug up and all that infrastructure exposed, he's seeing both the past and the future. The past is the chaotic accumulation of decades or centuries. The future is the attempt to rationalize it, map it, and make it locatable with centimetric precision.
That attempt is impressive when you look at what's involved. Let me walk through what happens when a utility is going to install a new high-voltage underground cable in a city today. First, they do a desktop study — collect all existing records from every utility, every municipal department, every archaeological survey. That's the Level D S. Then they do a surface survey — walk the route, mark visible features, look for signs of underground infrastructure like manhole covers, transformer vaults, vent pipes. That's Level C.
Then the geophysics?
Then they bring in G. , electromagnetic locators, sometimes acoustic sensors. They scan the entire route. data is processed with software that can differentiate between different types of buried objects based on their reflection signatures. A water pipe and an electrical conduit look different to G. because of differences in material density and dielectric constant. The result is a subsurface map with horizontal accuracy down to about two or three centimeters.
That's the centimetric precision Daniel asked about.
That's the starting point. Then, at critical locations — especially where the new cable will cross existing utilities — they do vacuum excavation to physically expose what's down there. They verify the depth, the exact horizontal position, the condition of the existing utility. Then they design the new installation to maintain proper separation. The whole process might take weeks for a single city block, and it can cost tens of thousands of dollars before a single meter of new cable is installed.
That's for a competent utility in a well-managed city. I imagine there are plenty of places where the process is more like: "I think the gas line is around here somewhere, let's dig carefully.
And that's where the damage statistics come from. The Common Ground Alliance data I mentioned — four hundred thousand damages per year in the U. — those are mostly from smaller-scale excavation where the full S. process isn't followed. A contractor putting in a fence post, a homeowner planting a tree, a telecom crew trenching for fiber. They either don't call for locates, or the locates they get are inaccurate.
The fiber expansion is probably creating a whole new wave of this.
It absolutely is. The broadband build-out has meant millions of new underground installations, many of them in residential areas with poor existing records. There have been numerous incidents of fiber installers hitting gas lines, including some that caused explosions. It's a reminder that subsurface precision isn't just about high-voltage cables — it's about every utility that shares the underground space.
Let's talk about what the future looks like. If we project forward twenty or thirty years, does the overhead wire become a thing of the past in cities?
In dense urban cores in developed countries, probably yes. The cost is high, but the reliability benefits, the safety improvements, and the aesthetic pressure all point in one direction. In suburbs and rural areas, overhead lines will persist for a long time — the economics just don't work otherwise. And in developing countries, overhead lines are still the default because they're faster and cheaper to deploy, and electrification speed often matters more than aesthetics or storm resilience.
We're heading toward a world where the urban-rural infrastructure divide becomes even starker. City dwellers walk on pristine streets with everything hidden below. Rural areas keep the poles and wires.
There's also an interesting technological development that might change the equation: high-temperature superconductors. If we ever get room-temperature superconductors — and I should emphasize that's still very much in the research phase — the economics of underground transmission change dramatically. Superconducting cables can carry vastly more power in a much smaller cross-section, which means smaller trenches, less material, lower installation cost.
Room-temperature superconductors have been "ten years away" for about forty years.
But the progress in the last decade has been genuine. The current record for superconductivity at ambient pressure is around minus twenty-three degrees Celsius, which is still cold but doesn't require liquid nitrogen. If that creeps up to temperatures that are practical without extreme cooling, it would be transformative.
Until then, we're working with cross-linked polyethylene and careful digging.
tags and three-dimensional G. and vacuum excavation. It's not one technology that gives us centimetric precision — it's a whole stack of technologies, plus the organizational discipline to use them properly.
I want to pick up on something you said earlier about the equity dimension. If undergrounding improves property values and reliability, and it's concentrated in wealthier areas, that's a policy problem. Are any cities addressing this head-on?
Some are trying. In California, the Public Utilities Commission has been pushing utilities to prioritize undergrounding in disadvantaged communities, not just in wealthy areas where the property-value math is easiest. The argument is that low-income communities often have the worst reliability because their overhead infrastructure is older and less maintained, and they're least able to afford backup generators or hotel stays during extended outages. So resilience investment should be directed where the need is greatest.
That makes sense from an equity standpoint, but I imagine it runs into the same cost-per-customer problem you mentioned.
It does, and it requires either cross-subsidization from the broader rate base or direct government funding. Neither is politically easy. But there's a precedent in rural electrification — in the nineteen thirties and forties, the U. government heavily subsidized extending power lines to rural areas that private utilities considered uneconomical. The argument was that electricity was a public good, not just a market commodity. The same argument could be made for undergrounding in underserved urban areas today.
That historical parallel is interesting. The Rural Electrification Act basically said: the market won't do this, so the government will. And it worked.
It transformed the American economy. By nineteen forty-five, rural electrification had gone from about ten percent to over forty percent, and by the nineteen fifties it was above ninety percent. That's the scale of infrastructure transformation that's possible with sustained public investment.
Do you think we'll see anything like that for undergrounding?
Not at that scale, because the problem is different. Rural electrification was about extending access where none existed. Undergrounding is about upgrading infrastructure that already works — it's a quality improvement, not an access expansion. That makes it politically harder to justify massive public spending. But I think we'll see targeted programs in specific contexts — post-disaster rebuilds, historic preservation districts, high-density corridors where the economics are strongest.
Daniel's question about Jerusalem is interesting in this light because Jerusalem doesn't fit neatly into any of these categories. It's not a wealthy Western European capital with unlimited infrastructure budgets. It's not a developing-world city where overhead is the only option. It's a historically dense, archaeologically complex, symbolically charged city that's trying to modernize while preserving its character.
It's doing so in a security environment that adds another layer of complexity. Underground infrastructure in Israel has to be hardened against potential threats in ways that most cities don't consider. But that's probably veering into classified territory, so let me steer back to the civilian engineering.
So let's summarize where we've landed on Daniel's questions. First, the precision: centimetric accuracy in placing high-voltage underground cables is achieved through a layered process — records research, surface survey, geophysical scanning with G. and electromagnetic locators, physical verification through vacuum excavation, and increasingly, embedded R. tags and tracer wires for future locatability. It's not one magic device. It's a disciplined methodology.
That methodology is formalized in Subsurface Utility Engineering standards that are increasingly adopted worldwide, though implementation quality varies enormously.
Second, the prevalence and transition: underground power distribution is the norm in dense urban cores in developed countries, but it's far from universal. The cost differential is enormous — five to ten times more expensive than overhead — and the economics only work in high-density areas or where reliability and aesthetic benefits justify the premium. We're moving toward cleaner streetscapes, but the transition will take decades and will be uneven.
Some of those overhead wires Daniel sees are indeed legacy telecom, which is a separate but parallel undergrounding challenge. The poles often carry both power and communications, so you can't underground one without dealing with the other.
The picture that emerges is of a slow, expensive, technically sophisticated process that most people never think about because when it's done well, it's invisible.
That's infrastructure in a nutshell. The best infrastructure is the kind you don't notice. You flip a switch, the light comes on. You don't think about the hundred and sixty-one kilovolt cable buried under three feet of soil and concrete, the substation that stepped the voltage down, the distribution transformer on the pole or in the underground vault, the service drop to your building. It all just works.
Until someone digs in the wrong spot.
Then it very dramatically doesn't work. Which is why the precision matters.
Now: Hilbert's daily fun fact.
Hilbert: The world's largest known underground lake is the Lost Sea in Sweetwater, Tennessee. It covers about four and a half acres and is part of a vast cave system, and it contains a species of blind fish found nowhere else on Earth.
Blind fish in Tennessee.
That was entirely unrelated to anything we discussed. I appreciate the consistency.
Here's a forward-looking thought. The infrastructure beneath our feet is becoming more crowded, more critical, and more precisely mapped. But it's also becoming more vulnerable — to cyber threats, to physical attacks, to the simple accumulated errors of a century of incomplete record-keeping. The challenge of the next few decades isn't just installing new cables with centimetric precision. It's creating the institutional and digital infrastructure to maintain that precision across generations.
That's a coordination problem, not an engineering problem. The engineering is hard but solvable. Getting a dozen utilities, three levels of government, and a century's worth of legacy records to all align — that's the real challenge. Thanks to Hilbert Flumingtop for producing, as always. This has been My Weird Prompts. You can find every episode at myweirdprompts.We'll be back next week.
