Central Subway y u no make sense?

I really have no idea what's going on here.

So first they dug these narrow, deep holes and dropped girders in.

Then they dug thin trenches between them and poured walls.

Then they excavated out the dirt between those walls! What?

This city is under constant construction, so I've seen a lot of basements and foundations being dug, and I've never seen it done in that order before. They always dig a hole, shore up the sides as they dig, and then pour the walls. This seems like way more effort.

So next, they dropped girders on top of the hole, at street level, presumably to hold up the street that's going back on top -- except they only left like 4' underneath the girders, so what's going under there? Why did they do this at all! I thought the point of digging this hole was to build the head-end for the drilling machine, meaning the receiver for a bunch of conveyor belts trailing behind a chewing disc that is around 20' across.

What the fuck are they doing? Their blog is pretty good with the updates about what they're doing right at this moment -- "we're pouring walls" -- but there's absolutely no big-picture overview that I've found that explains what specifically they're building or in what order. What's the thing under the freeway going to look like? Where exactly does the drill go in? Are they manufacturing tiles on site or shipping those in? The animation was awesome, but not specific to this project.

The plan set for this project must be thousands of pages, and I suspect that the first 4 pages would explain everything I want to know, but I don't know where to find it.

And most importantly, once that drill is in the tunnel, how do I get a tour?

Related: I still have not been on top of the new Bay Bridge span. Hook me up.

Tags: , , ,

22 Responses:

  1. djm says:

    Maybe the soil is too friable for usual ground-wall pinning techniques?

  2. Logan Bowers says:

    Often times they will dig holes inject "grout" into the ground to stabilize it. This will prevent the soil from collapsing and causing sink holes or slides while they dig a bigger hole or bore the tunnel underneath.

    Soil in Seattle, for example, was unexpectedly sandy when they bored the Beacon Hill tunnel for light rail. They monitor the amount of soil extracted per unit of movement to make sure the tunnel ceiling isn't collapsing in and getting sucked up by the cutter head, but even still they managed to open some sink holes under homes 200ft above (which then had to be eminent-domained).

    They may be trying to protect what looks like an elevated freeway structure, or they may be hardening up the soil so that it's easier to excavate predictably.

    • jwz says:

      I also can generate theories like that, but I wish someone would just tell me what's actually true!

      Surely all the land downtown, east of First street -- which is literally sand on top of broken ships in the former bay -- is less stable than 4th and Bryant, which had always been actual land, and yet, this is not how they dig basements for tall buildings over there.

    • Francisco says:

      This is probably the right answer. Jet grouting is very common in subway construction since subway lines transverse various types of soil and rock with varying degrees of degradation. Excavating argilous material (clay), particulary organic argilous material can be time consuming and ineffectual. Because vary small grain sizes and high water retention this soil behaves in a way somewhat analogous to digging in a mix of cornstarch and water, when you take out a handful it just behaves like a very viscous liquid and fills the hole back in. It's therefore more cost effective to build a wall with jet grouting which very much like regular holes are drilled except when the drilling head is pulled back out it comes out pumping cement and water and mixing it with soil to form a 'soilcrete' column. These columns are placed juxtaposed and form a wall, that is therefore excavated. Since these columns are not reinforced and concrete by itself has very little resistance to tension, the girders could also possibly serve the function of supporting the walls after the excavation making the bending moments smaller (albeit judging by their sizes it's not their only function). This material is also horrible to build foundations on and takes years to stabilize when compared to other types of soil (which can lead to buildings thats sink more than a meter after constructions, in the worst cases some sides more than others). If the soil layer is very thick or depending on the building's load, jet grouting is also used to create a foundation where the load is transferred to a layer of rock or sand tens of meters below the argilous material. You can google 'jet grouting' and a bunch of stuff will come up. I live in Rio where this type of soil is common and these techniques are widely employed here. Mexico City is the largest city in the world and was built basically on top of a swamp (where the soil usually has a large content of clay and silt).

      This can be easily found out asking someone on site :)

  3. Jeff says:

    If you are especially close to the water table, then digging down isn't going to create a hole, or at least it won't be a hole when you come back in the morning, it will be a pond.

    Poring concrete walls before excavating the inside will make this water inflow to a manageable, level. Concrete isn't water proof, but it doesn't allow for any flow of water. And when concrete gets wet it doesn't turn into entirely structurally useless mud.

    Also, depending on the geology (and nearby existing structures), the walls might also be necessary to maintain the general structure of the area. If you are diging through solid rock, you can go straight down even if you are 10' from a structure. If you are on mud, garbage, shipwrecks and what have you, you can't do that.

    There is probably some specific cost benefit analysis. Doing it in this order would have some up front costs, sure, but this might be offset by the site being forever then geotechnically sound. Work slowly and carefully once, then go crazy.

    Do they have buildings in SF that need foundations 50' deep? That are right beside existing big buildings?

  4. Yeah, the soil there is basically mud. They had to pour walls first, then dig out the hole. Remember how the Claw was digging in a pool of liquid? That's wasn't some special drilling "mud", it was just mud.

    I'm guessing the current 'dig out four feet then stop and put in girders' phase is just because it's easier to put in the girders that way, just carry them in with that excavator. Digging out the rest of the dirt later shouldn't be a problem, they don't have to do it from the top.

    • jwz says:

      They were definitely pouring something into the hole as the claw was digging, so that wasn't only ground water seeping up.

      Also, like I said: downtown. I've never seen them do it this way there, and that's actually in the old bay. So why here and not there?

        • Grey Hodge says:

          This is the reason why they did it. These types of construction make very solid, waterproof walls. The bathtub around the WTC foundation was the only part of the entire thing that survived the collapse. As for why they're doing it here adn not on other buildigns is because this is a project that is huge in size, crossing various types of ground, and must be as watertight and long lasting as possible. Building basements are much smaller, and most importantly have much less stress in the way of load. Buildings are static load, but this will be a dynamic load and probably have more dynamic loads above it (streets). Also, this isn't actually more work, you're using the trench to your benefit, rather than trying to create a temporary restraining wall and a vertical concrete form.

        • Charles Stephens says:

          Oh look, that page has that machine again.

    • Jeff says:

      With vertical walls, you are either going to have to anchor them back behind them, or hold them apart, top and bottom.

  5. They used this same approach in the 70' deep excavation for the new trans bay terminal.

  6. George McKie says:

    If I have understood your description of the timeline, and this page correctly: they dug narrow, deep holes to hold the decking support footings; then they constructed the excavation shoring walls with slurry trench methods; now they are placing the decking trusses on the footings. Presumably, the four foot gap provides enough headroom to excavate after the decking is in place.

    However, this is all supposition.

    • BrianVan says:

      Given the prior photos, you're probably correct. Same construction method used to create the hollow foundation "bathtub" of the WTC in New York, since the land quality in that spot is similar to SF's "sand on top of broken ships". Again, it's impossible to get a vertical wall down to bedrock (70ft) when a massive saltwater estuary is flowing through your site perimeter.

  7. MattyJ says:

    Why the hell are these charts so small? From one of the construction companies in the joint venture. It describes the type of ground they'll be boring through but I REALLY want to see those charts:


    This site indicates that they're currently building the launch box and briefly describes what they're doing, alluding to it being done this way because of the proximity of existing infrastructure:


    They did a similar thing when excavating the first tunnel for the new Doyle drive. It allowed them to dig that tunnel out literally two feet from the existing roadway. They had no room to dig wider, build the wall, then back fill it.

    • jwz says:

      Someone sent me a link to this early, out-of-date Environmental Impact Report. I guess what they're actually building is "Alternative 3B" or something a lot like it, starting on page 2-36. There's not a whole lot of detail there about construction, but there is this image:

      which shows at least one thing I was curious about: that the entire block under the freeway will be cut-and-cover, with actual boring beginning at Harrison, just before Whole Foods.

      Looks like they've got around fifty more feet of their dirt in boss's hole!

      I wonder what that "construction shaft" is.

      • anonymouse says:

        I was just down there, and it's pretty clear what they're actually doing now: the trenches were there to build the walls all the way down to the eventual tunnel floor level, and the walls are there to support the massive cross-beams you see, which will actually support a temporary roadway, so that they can reopen the street to traffic as they dig the rest of the way down and then build the cut and cover bit of the tunnel in the hole. I've seen something very similar done in LA when they were building the subway there: expect a bumpy but reopened Fourth Street in a month or two, which will stay that way for a couple of years at the least.

  8. Anthony says:

    The plans are top-secret, and anyone wanting a set must sign in blood that they will on no account allow terrorists or unqualified contractors get their hands on a copy.

    Even though pretty much every single CalTrans project has plans available to anyone willing to download hundreds of MB of PDFs, even for project which might actually be security sensitive, like the new Bay Bridge main span.

  9. Anthony says:

    The big picture won't be available in the first few pages of the plans, though there may be somewhere an "executive summary" describing what's going on. In this particular case, the holes with the girders are soldier beams for the shoring. The thin trenches connecting the girders are either a slurry wall or a more permanent retaining structure for the excavation. They dig down a few feet, then install the first level of walers, which are the cross-beams which transfer the loads on the shoring across to the other side (and vice versa), so that the vertical soldier beams don't have to be really f***ing huge. There's only 4 feet underneath them because doing the connections to the excavation walls is a lot easier if the guys can stand on the ground while they're doing connections, and not have to bend over too much, either. Once all the welding or bolting is done, they'll dig down a ways until they put in the next level of walers, and repeat the process. Or until they reach the bottom.

    Basically, there are three ways to support an excavation against sideways earth pressure:
    1. Really big vertical beams to resist bending. If the excavation isn't too deep, or the soil is reasonably cooperative (is rock, basically), this may be enough.
    2. Walers across the structure so that the loads in each direction balance each other out. This works if the far side isn't too far away.
    3. Tiebacks into the ground behind the excavation wall. This leaves you lots more clear space to work than do walers, but then you have to know what your tiebacks are going into. Tiebacks can be assumed to go at least as far horizontally as the excavation is deep.

    There's a lot of stuff buried in the streets in San Francisco, and if you're too close to the sidewalks, lots of basements and foundations you don't want to mess with. So on this job, tiebacks are right out.

  10. Josh says:

    When they do decide to let you onto the bridge as part of a school field trip, you bring the camera, I'll watch the kids: if my kid can come too. I've been trying to figure out how to get him to be able to check it out early, too.