Comment by epistasis | Hacker News Reader

Comment by epistasis | Hacker News Reader

I'm very grateful for those posting and upvoting interesting construction articles on HN, this is a key area of our economy that is stagnant and could really use some improvement. Traditional media do a terrible job of coverage, and spreading articles like this helps get better mainstream coverage.

A lot of the bad things about construction come down to incumbent players being able to force extremely costly methods on anybody who wants to improve the world. So a lot of the challenge is political. Solving political problems is really hard, but when technology can sidestep the problem, there's a lot that becomes possible.

For those in the Bay Area, we are seeing aversion to cut and cover be a huge hassle for our stations. The depth of some of the new transit stations will be a huge inconvenience for generations of transit users, all to satisfy a short term disruption of a far smaller number of people. This problem could have been solved with money, for those affected, if it were not for the politics.

Deep stations are also unbelievably expensive. Deep tunnels are not that bad, but the resulting stations explode the budget for transit to the point getting the project green lit in the first place is nearly impossible. Plus with the same set of dollars you are building a lot less transit.

The deepest station in Seattle has smelled like someone tried to make cheese from moldy gym socks since the day it opened.

I believe it opened and immediately shut again so they could try to deal with the water problem [though responder says no]. I went through when it was briefly open and again after. They made it less bad, but it was still pretty awful.

A club I belonged to met at a library near the station so it was easiest for me to just hop on the light rail after work and then take it back to the bus terminal after but gods did I want to hold my breath the entire time.

> The deepest station in Seattle has smelled like someone tried to make cheese from moldy gym socks since the day it opened.

Maybe it's just a matter of opinion, but I never found the Beacon Hill light rail station to smell like anything other than raw concrete, similar to DC metro or any other deep tunnel system.

I lived in the neighborhood at the time, and don't remember the station shutting down due to water ingress. Opening was delayed due to issues with overexcavation, but that threat was with sinkholes opening to the public above, not issues with the tunnel itself.

My nose is pretty good at picking up mildew smells. I've been in plenty of rooms where half the people can't smell anything.

After years of rhinitis my sense of smell is half-shot, but I can still walk into a room and find the forgotten towel moldering away under a chair or table. Why couldn't I have lost that power and kept something more pleasant?

Deep stations also add a bunch of hassle for the users. It just takes longer to get down 5 stories underground.

It's absurd to talk about transit times to destinations when just getting to track level and back can take longer than the trip

This is one of those things that people overlook. Travel times should be "door to door" or some similar comparison; sure it makes cars look a bit better but it can make things like trains and walking look even better. A two hour flight might be bounded by 1+ hours on either end (including security, etc), whereas a two hour train ride might literally be two hours + five minutes walk on each end.

You need to factor in the fact that adjoining landowners are often rich campaign contributors, whereas the patrons of public transit are often the poors or the invisible help.

As is often the case, the answer is "follow the money".

Exactly. When I get to a cut and cover station, it's often just a single flight of stairs down. It's often only 15 seconds or so to reach the track and wait for the train.

Stations along bored tunnels sometimes take as much as 5 minutes to go from street to train. One I used to regularly use had three separate multi-story escalators, in addition to long walkways between them.

So even if cut and cover disrupts businesses and car traffic, I can't help but wonder if decades of extra delays to reach the tracks from many thousands of people a day doesn't wind up outweighing that.

Is there a reason bored tunnels need to be substantially deeper? Does boring lead to structural stability issues that the installation of the tunnel structure doesn't mitigate?

I do know that tunnel boring machines can disturb the surface, and that substantial effort goes into minimizing this. (That is, minimal surface disturbance isn't automatic on account of the fact that you're boring.) I believe it sometimes has to do with how much pressure they need to apply to the tunnel face; the ideal pressure for cutting can lead to sink or swell on the surface. Presumably this is less of an issue the deeper you go. I don't know much about this stuff though. This is maybe relevant:

> This thesis summarizes and evaluates the performance of Tunnel Boring Machine (TBM) construction in Singapore's MRT network. Surface settlement induced by the tunneling process can cause damage to underground utilities and foundations and buildings and/or disrupt daily life by damaging roads and pavements, and is used in this thesis as a measure of performance.

https://dspace.mit.edu/handle/1721.1/82809

So I did a little research on this, and it seems to be two main things.

Partly yes it seems to be a question of stability -- you're not going to bore and leave just a couple of feet of soil above, because of the way you want the arch of the tunnel to support itself, and that's not going to be strong enough. And close to the surface you've got looser soil etc. Going deeper does provide greater stability, in a way that isn't needed for cut-and-cover because that is supported by steel beams, not arches. But this doesn't generally require you to go crazy deep.

The bigger factor behind these crazy-deep tunnels seems to be pre-existing infrastructure. If you're doing cut-and-cover, you're digging out all of the infrastructure first and then replacing it -- it's highly disruptive. While if you're boring a tunnel, you need to go below everything that already exists -- sewers, water lines, deep building foundations, pre-existing subway lines that it crosses underneath, etc.

> The bigger factor behind these crazy-deep tunnels seems to be pre-existing infrastructure. If you're doing cut-and-cover, you're digging out all of the infrastructure first and then replacing it -- it's highly disruptive. While if you're boring a tunnel, you need to go below everything that already exists -- sewers, water lines, deep building foundations, pre-existing subway lines that it crosses underneath, etc.

Makes perfect sense!

And yeah, it's yet another case of path dependence: we'd build far different cities if we had the option of designing "from scratch" rather than layering retrofit over retrofit under retrofit.

In an ideal world, we'd have tunnels connecting B1/B2 of every major building in a city, and many people's trips would be elevator-transit-elevator.

Why are deep stations expensive but tunnels cheap? I'd guess the opposite, tunnels being larger.

In an urban subway line, stations may cover ~25% of the line. The platforms are long, while the distances between stations are often short. And while tunnels are standardized and their construction is largely automated, each station is a unique project that needs to be adapted to the specific location.

Because you need to dig out all that dirt, and this is normally done cut and cover. Even if you are only digging for stairs and elevators (building the station in the tunnel as Madrid does) that is a long way down. For shallow stations you can tell most people to walk down the stairs, but for deep stations nobody will so you have to build enough expensive elevators to handle everyone (and you still need the stairs for fire escape). Deep stations also mean people will be on the elevator longer so you need even more of them. (people may be running late so they need to load fast and it better not be long for the elevator to come when someone needs to get someplace)

> elevators

How are the elevators any more of a problem than in buildings, where they are commonplace, often for much taller structures (i.e., many more floors), and not considered a major expense?

Edit: Thanks, good points about the volume/throughput.

Metro stations have enormous throughput in comparison to a building.

A subway station needs to be able to handle, during rush hour, bursts of 100 people arriving every 2-3 minutes. A building generally does not.

This is why even deep subway stations usually go with multiple levels of escalators to keep the flow of people moving continuously.

Escalators usually carry more people if they are offered at least. Elevators get disgusting fast.

For deep stations they don't work as well - you end up with a lot of switch backs.

Depends but you can have very long ones, too. In koreatown LA runs one of the longest escalators in the world down into the depths. Even the escalators at Universal Studios Hollywood have only a few landings and get you down an entire mountainside of vertical drop. I’d still take a switchback over cramming into what is basically a dumpster and urinal.

In the SF Bay Area, BART closed the station restrooms and so now homeless people and drunks use the elevators and escalators as toilets

https://www.sfgate.com/bayarea/article/Rights-group-sues-BAR...

Could be of note that Stockholm is building the world's second-deepest subway station right now[1]. Though there's no real alternative to that because it needs to go under water (well, in the ground under water) towards the next station.

[1]: https://www.youtube.com/watch?v=TEYvywnJy78 (Swedish)

When there isn't really a choice, fine build a deep station. Montreal is building a station 72m down, but it's driven by surrounding geography. What's a bad decision is deep tunnelling just because, like Toronto on the York subway extension. Deep bore under suburbia is really wasteful spending and that same transit budget could be building a lot more transit.

https://en.wikipedia.org/wiki/Édouard-Montpetit_station#REM_...

I lost 25 lbs in 6 months running down and walking up the looooong escalators in Stockholm because it took way, way too long to stand and ride.

Here's my startup thesis for this space, free for anyone who wants it

1. Most delays in tunneling projects are due to elongated decision making processes 2. The majority of the cost of the machines are their initial construction, not operation or maintenance

Solution: fleet of TBMS are set off in arbitrary directions that bounce off the city borders like the screen saver with the bouncing ball. This creates a large network of tunnels all across the city. When it comes time to decide to put new infrastructure in, one of these tunnels is already there somewhere close, just have to dig down to it. 0 decision fatigue/cost, max efficiency for cost per tunnel foot.

How could this problem be ‘solved with money, if it were not for the politics’ ?

Directly paying everyone living/working nearby to get them to accept cut and cover construction? Wouldn’t that cost even more then building it deep underground, after all the legal fees and court challenges, once it’s said and done?

It would depend on the implementation. Cut and cover could be done in blocks rather than shutting down long corridors at a time, combined with legislation setting clear limits on affected compensation and their standing to sue over small delays. That could easily pencil out to much cheaper than deep bored tunnels and stations.

Legislation that restricts local resident’s standing to sue? Can you provide some examples where that has ever been implemented in the US, in any scenario, or even judicial opinions to that effect?

NYC is paying $4B/mile for their latest tunneling projects. That works out to $200M/block.

Lot of money to go around, I'd argue its more political will than cost.

Don't forget theres years of legal fees & court challenges even with deep tunneling!

Alon Levy's Transit Construction Costs project breaks down where NYC's (extremely) high costs go - https://pedestrianobservations.com/2023/02/06/our-constructi...

There's no simple reason - basically everything that can cost more, does.

NYT had some great articles too.

A lot of it is that every last bit of expertise is outsourced.. not just physical labor and renting equipment but even the project planning, management and oversight. You'd think given decades of capital programs lined up, the state would want to have their own planning/management/oversight staff so that interests are more aligned.

You can't pay yet another vendor to care about & enforce cost discipline upon all the other vendors you are paying..

You also can't discount antiquated union work rules & strong lobbying. NYC transit work sites are more heavily overstaffed compared even to the equivalent Paris transit project. "Our unions have more power than in France" is a pretty high bar. From having 3x the staffing on the Tunnel Boring Machine to minimum required union machine oilers on-site based on the level of hand oiling that 1970s equipment needed (far more).

It's like if IT's staffing / roles were frozen in time from the 1970s and we still paid someone to sort punchcards.

> NYC is paying $4B/mile for their latest tunneling projects.

So only what, about 30x as much as Madrid spent?

This problem couldn't, but a lot of others could. Grandparent comment spoke generally about all construction.

Can be solved by doing an emergency nationalisation, i.e. pay first, everything else after, which leaves challenging the compensation to the after.

Reply on news.ycombinator.com