Enjoy this engineering analysis and speculation about San Francisco's ever more leaning Millennium Tower

Originally published at: Enjoy this engineering analysis and speculation about San Francisco's ever more leaning Millennium Tower | Boing Boing


It’s a little weird when they’re using terms that are valid in both construction and medicine. :grimacing:


16 minutes! Whoo-wee. Let’s summarize:

The Millennium Tower was built on a 10 foot thick foundation slab poured on top of a grid of 70 foot long “friction pilings”. These pilings rely on the friction between the skin of the piling and the soil to transfer much of the load of the building’s weight to the soil around the pilings. Imagine this force extends out around the sides of the piling diagonally and downward. The surface area of the bottom of the piling carries the rest of the load. At the bottom of the piling the remaining load is carried away in a conical shape downwards.

When they sank a test piling and measured it, they had an estimate for what kind of load one piling could support.

Friction pilings are at their maximum effectiveness when they’re spaced far enough apart; otherwise the load is shared between neighboring pilings, diminishing each piling’s capacity to carry its load.

In this building they overcrowded so many pilings into the foundation space that they left gaps less than two-people-wide between them. These factors combined to turn the grid of pilings into something that acts less like a grid, and more like an 80 foot thick foundation slab, pushing more of the load straight down.

The pilings ended at the boundary layer between a sand layer and the Old Bay Clay layer. Because the load forces at the bottom of each piling spread out in a conical shape, and the pilings being so close together, their cones overlapped. This concentrates more of the load forces under the center of the slab, and less to the sides. The load now looks almost like a hemisphere, with the highest forces in the middle.

Next, using easily obtained numbers, the load on the bottom of one piling can be calculated to be around 15,000 pounds per square foot. This is nearly the capacity of the Old Bay Clay just below it. But remember, the load is now concentrated at the center of the slab, where it is likely exceeding that clay’s capacity, resulting in plastic deformation of the overloaded clay in the middle of the footprint.

These assumptions are supported by evidence that the foundation slab has begun to “dish”, sinking more in the middle of the building than the edges (he said it’s actually more to one side.)

If you want a mental picture, imagine the foundation grid of the building is perched on a giant soft, squishy ball.

Bedrock pilings would have been the other choice. These are pilings that would have been sunk through the Old Bay Clay layer all the way down to the stable Franciscan Bedrock layer below it.

He then absolved the developers of blame for choosing friction pilings over bedrock pilings. If the engineers came to them and said “You have two choices that could work; friction pilings, or for an extra $4 million you could have bedrock pilings. We recommend bedrock pilings.” what do you think the developer would do? Any person would hear “either option works, so pick the cheapest.”

And that was it – after 16 minutes.


As a San Francisco engineer, when I learned about this, a fairly obvious truth occurred to me: Regardless of how we got here, if the pile - to - earth connections are shifting, the pile - to - building length should also be able to shift in compensation. One solution is to insert hydraulic pistons into each. The forces involved are quite practical. I posted a more complete discussion of the issue, this solution and potential issues (and solutions) at


That’s like ‘put a penny under the table leg that is too short’ only bigger.

Is going down to bedrock better? I imagine it will shake so tying yourself to it rather than floating above it might make things worse. But it shouldn’t tilt much.


Dont be silly, the best solution is clearly a team of goats, four or five should do, some kind artisanal hemp rope for the green credentials and the will to just keep pulling. I recommend an olive tree on the other side of the street.


A tragic comedy of money, greed, & abject stupidity. The Netflix series should be entertaining.


I propose one floor be converted into a huge magnet to pull toward the buildings opposite the lean. That should make things interesting for folks in the neighboring buildings.


All there cutlery will have to be swapped out for plastic.


baby bar GIF


Thank you very much for the Readers’ Digest Condensed version. I wish more of the videos had a synopsis like that.


He also said the reasoning behind saying “no” was the developer felt the local government or the neighboring project would pay the $4 mil if they claimed it was only needed for the other party’s safety and not their own. They were wrong.

Basically, they built their piles too close together, and are too close to other buildings without the proper engineering.


Artist’s conception of the developer’s new proposed mitigation plan.


In all my years of doing machinery design. There are certain places in equipment where you can definitely go with less expensive options and the customer won’t care. There are other places that if I upgrade and and use a slightly more expensive part, or a more expensive part. I can make that part of the machine run and last much longer before needing maintenance or replacement of the part. Bedrock pilings are like that. Despite the cost for $4,000,000 for putting in pilings to the bedrock. The guys who gave the developers the option for the friction pilings were idiots in my opinion. Bedrock pilings are tried and true and hold up your building.


Especially in a ‘liquefication zone’


Especially that! Nothing like watching your building sink into the ground quickly.


So I built a second Millennium tower. That sank into the clay. So I built a third tower. That burned down, fell over, then sank into the clay. But the fourth one stayed up.


Engineers in California know how to build shock absorbers under big buildings

OTOH everything I’ve ever read about earthquakes said the worst place to be was on top of something soft, like landfill


i think the first one was clay

note also kind of leaning ( and didn’t last very long. )