More American Station Construction Extravaganza
Los Angeles has plans to extend its urban rail network. They’re taking forever, because construction costs are extremely high – and Nick Andert just pointed out one reason: the station caverns are huge. The overage on the proposed northern extension of the K Line, also known as the Crenshaw/LAX Line, is a good deal larger than on Second Avenue Subway, making it the most wasteful station construction that I am aware of. This is driving up the construction cost estimate to, depending on which alignment the K Line is to take as it goes north into West Hollywood, around $1 billion per kilometer.
Nick provides some station footprint diagrams. The K Line’s stations are designed for a maximum train length of three cars, or about 81 meters in total. The stations on the proposed extension, however, start at 124 meters when there is no crossover, or 50% overage, and most are 300 meters with crossovers. In other words, an underground light rail extension with trains less than half the length of New York City Subway trains is proposed to have stations about as long as those on Second Avenue Subway, which are already about twice as long as they need to be by New York standards. (In the low- and medium-cost countries for which I have this information, the overage is not 50%, but ranges between 3% and 17%.)
The 50% overage without crossovers is completely unjustifiable. But the crossovers, which turn the 50% into nearly 300% overage, are equally unjustifiable. It is not normal to build bored tunnel subways with so many crossovers, precisely because it’s expensive to blast caverns for them. Marco gives the example of Milan M4, which, counting the soon-to-open extension, has four crossovers in 15.5 km and 21 stations.
To this I can add that the Copenhagen Metro, built with bored tunnel with blasted caverns for crossover, has on the original line just two underground crossovers; the City Circle Line has only two as well, plus two on the M4 branch. There are more crossovers above ground, where it’s not so expensive to build them, but still less than one per station. This is a system designed for 24/7 operation; crossovers are required to allow trains to run on a single bidirectional track at night to permit maintenance, one track at a time. Without this constraint, even fewer crossovers are needed – only at the ends of the line, which includes the end of each operating segment if the line opens in stages.
If the K Line extension’s split between station and tunnel costs is similar to that of Second Avenue Subway Phase 1, 3:1, then shrinking the stations to the normal overage of a few percent would slash their cost by a factor of close to four, which would cut the line’s cost by more than half. This is what the extravaganza of crossovers is doing to Los Angeles and its ability to build mass transit infrastructure.
Pedestrian Observations 
LA metro must have plenty of money because everything does look so big and impressive, and often within view of the jammed highly engineered freeway system, probably the world’s most expensive.
Those look more like potential easement drawings rather than actual designs. Easements are always larger than what is gong to be built. So much will depend on whether these are mined or cut and cover stations.
Also I’m not sure you are correct on Copenhagen metro, the tunnels were constructed using EPB TBM’s, most of the stations were top down cut and cover, so I’m not sure there was much if any blasting given the ground conditions under Copenhagen.
Difficult to argue you need more than 2 crossovers, one at the D line and one at the end
I’m not arguing anything, just saying these look more like easements than designs.
Yeah but why ask for such generous easements?
hedging their bets???
Hedging bets? Hardly. Just the usual “aiming for the moon.” It always works. They always hit their target of maximum expense and minimum utility. Always. USA USA USA Olympic rent-seeking gold medallists 500 times in a row, sweeping every discipline at every distance. USA USA USA!
As I understand it, Copenhagen built cut-and-cover stations but blasted caverns for the two crossovers? But admittedly I didn’t do as deep a dive there as in Stockholm.
Roadheaders and sprayed concrete. Either way at least three crossovers were mined rather than open cut
This is even more extravagant given the disappointing ridership of the Southern half of the Crenshaw line, 3170 per week day in April 2024.
What a million rides a year? Jesus that is shit for a whole line.
It is not a whole line. They opened it in sections, and the current temporary southern end is not a popular destination, nor did they make a direct bus from there to the airport. There is a bus paralleling almost all of its 5 mile route, and for transfers to the C line, just taking a bus on Crenshaw would be faster.
Yeah, there is approximately a 2.25 mile gap between its current end and its future end, because the station in the middle is a transfer to the LAX people mover, which has been delayed and is holding up this last segment of the K line.
In the EIR, LA Metro states it would like to top-down cut-and-cover most of the stations. I presume this includes the crossover box, which perhaps isn’t as bad as mining. I wonder what was stopping the tunnels from being cut-and-cover, too, if they’re willing to dig up that much of the street in the first place
Cut-and-cover stations at this depth should still minimize the overall size of the dig.
Of course. The design here is bizarre.
Out of familiarity, the reference point I have for a line of this type is the Toronto Eglinton LRT. There, typical station box length was 122m for platforms a little over 90m. There are two underground crossovers; one midway on the line inducing a 270m station box. The second is a double crossover (with a pocket track) at the end of the continuous underground section that brings the length of the mined cavern of Laird station to over 490 metres; it’s used for scheduled short-turns. It’s really immense. Incidentally, the underground stations on the extension of the line are somewhat shorter than on the original segment, at a bit over 110 metres.
Why not turn the short run trains at Aga Khan Park & Museum on the surface?
We need a wall of shame for those responsible.
When a criminal is sentenced, the judge has a report of past offenses. Similarly, when planners, etc. are hired, their past offenses should be considered.
DOT should be tracking these offenses.
the counter to that is some people learn from their mistakes and so it isn’t fair to hold them punish them for something they wouldn’t do again. With a judge we know they did it again (well if found guilty) , but when hiring someone a for a future job we don’t know what they have learned and thus will do better next time. Indeed one thing that low cost cities have in common is they are always building and thus learning from the past.
Of course we do need to do something about the contractor who has figured out how to abuse the process. However we need to be careful not to punish honest mistakes too much. (I have no idea what is the case here)
Fair comment. I say, make the facts public and let those in charge of the money make the best choices.
Without this constraint, even fewer crossovers are needed – only at the ends of the line
This seems a bit restrictive. There should always be at least one crossover mid line so there is the option of running short turn service during planned major maintenance or unexpected issues. Ideally there would be two, one on either side of the CBD, so that a short turn service from either side can access the whole downtown (or if the maintenance/emergency is in the CBD, you can run service on either side and get people close enough to transfer to another service and still access the CBD).
In many cases short turns are planned (the Swiss Triangle: rolling stock, service and infrastructure should be planned together) in which case there is no need to add extra crossovers (another Triangle situation is if you do put a crossover either side of the CBD – it should be at a station with a connection to another line or a major bus layover to ease connections if required). Also as Alon notes if you are opening in phases the crossover at the end of each phase is plenty enough.
If your planning includes not just the normal operating pattern (with infra and rolling stock) but also a number of fallback/degraded (or just alternative) oppats (with their respective infra/equipment, to the extent it’s not shared with the default), and additionally an (at least notional) flowchart for switching between them, does that make it a tetrahedron? (“A stack of triangles, with successively decreasing size and likelihood.”)
You planning should include enough crossovers that you can close any single section of line for maintenance and still run 30 minute service on the whole line. This includes cross overs (which likely need more maintenance since there are moving parts). This implies enough separation between sections that you can have workers on one track when the other is in use.
Not many people will ride at 3am in most cities on most days, but you still want to be running service anyway then because the peace of mind that you could ride if something happens is useful.
3am you use a bus and you deal with the homeless so the buses are safe for normal people.
Go to youtube, and search for
VchLgXqia5k
It solves the bus issue.
Exactly. The point is to do cost-benefit comparison between tetrahedra. “We can save a lot of money on digging crossovers if we are willing to accept that overnight services will be bus rather than train. And 30 minutes headway as daytime fallback oppat is basically worthless for a metro, we might as well plan to use replacement bus service instead. Incidentally, we can also decrease costs by simply locating the two tracks next to each other, since during the regular nighttime maintenance windows, both can&will be shut down.”
I’d note separately that for surface-running rail (particularly for light rail) systems, it’s often a good tradeoff to buy a handful of portable crossovers for the system. In case of a multiple-day shutdown of a segment (riverside underplass flood risk, road reconstruction, etc.), the portable crossover can be transported to site and installed in a few hours, laid over the rails and joining them with ramps at each of the three ends. Perhaps 5 km/h limit, but it’s very short anyway. (Plus a wire to guide the pantograph.)
a bus is a useful fall back but then you need to staff them. Subways are easy to automate and run with minimal staff. This is important in high cost of labor areas. We are talking about overnight maintenance so I have strong doubts you will actually run that bus service in the real world while you will run the train if you can as the costs are so cheap if you design for it.
in general you also need to question if there will be a road to run the bus on. Often there isn’t because the subway crosses something where there is no bridge. Don’t forget to look 100 years in the future where the city has changed.
Not that I’m against a bus for this, but it isn’t my choice.
You always want buses anyway for other routes and need less in the middle of the night so you have spare.
24 hour running Friday and Saturday nights and night buses the rest of the week as in London and some other cities is a good tradeoff.
Basil Marte’s reply about cost-benefit is much better stated than I said. I endorse everything he said.
A bus itself is cheap. The cost is drivers for the bus, and not all of the cost is money. Good luck finding enough people to work a temporary overnight job driving as bus while you do maintenance. Drivers have a lot of training requirements, almost nobody will be keeping their qualifications up to date and there isn’t a secondary market for their skills (other than again companies who are paying their people). You can of course pull this off. However it will be additional management effort and cost.
Very often when building a subway you should be crossing geography that doesn’t have good road services. Very often cities will change their layout – while it doesn’t happen much over 10 years, over 50 years you will often see significant changes, a surface road today may not exist. A subway tunnel is likely to be in use for hundreds of years as it is much more difficult to change than a surface street. You need to thus ensure not only does your city have roads for the bus to use instead, but also that you have enough long term power to keep those roads around for your overnight buses.
Although there is no traffic overnight, can your buses provide acceptable backup service for a subway? Time sitting at every red light counts against you. If the roads are not 100% parallel to your line, time getting from the station to the road counts against you (of course if the important destination is not on the subway line the bus can potentially take a more direct route counting for them)
If you can make it work and provide acceptable service then a bus might be a good answer. However I question if you can.
London has night buses and heavy traffic and it is fine.
Night buses work.
Er, my point was exactly to partly disagree, henrymiller74. (I broadly agree with the things that Matthew Hutton said.)
The solution I was proposing is not a temporary night bus service to be an (approximate) 1:1 replacement of the metro only during maintenance. The whole point is that for most weeknights, entirely replace the daytime transit network with a nighttime network, a clean sheet design effectively independent of the former. Given the typical passenger densities overnight, the latter will probably be a timed-connection 30-minute bus takt. And, for obvious reasons, there will be a significant family resemblance between the two — if you do have any driverless metros, you may well include them; if a disproportionate share of nightlife just so happens to be found in the corridor of an extremely iconic tram line, perhaps you keep that running; and above all, even when you shut down the whole tram network (er, my city has an extensive one), all of which is drivered, as well as all drivered metros, both of these types of rail tend to coincide with the straight arterials that are the first-obvious-guess for where the night buses should run. It is not a problem if that coincidence is broken in some places, if the bus cannot follow the rail vehicle on the very same bridge because this time, the metro tunnels under at a place where no bridge exists; the night bus network can still get you everywhere in a reasonable time without being an 1:1 copy of the daytime network structure. It is in a certain sense a purely fortunate side effect of this arrangement, of always shutting down the drivered rail systems overnight, that multiple-hour maintenance windows are abundantly available.
As a separate issue, of looking decades (or even a hundred years) into the future. My overall impression is that over the previous century, heavily-trafficked surface roads have turned out to be more durable than most land uses, even if the character of the traffic on them has changed, both in its form as well as, implied by the land use changing, in its content.
Comparing our present position to a notional us moved to between WW1 and the Depression, what would (and would not) have been foreseeable back then, and thus what could we speculatively say now?
– It would have been possible to see motorization and road-ization coming. Mass car ownership and that lorries kill wagonload railfreight. Comparably, it should seem plausible that in a few decades, “drone” airmobility will have a value density threshold and safety sufficient to take over a significant share of small deliveries and passenger traffic. How that will come about is treated as an unknown for the purposes of this crystal-ball-gazing session, just as the specific technologies that made cars and lorries cheap and performant enough to take over would not have been known pre-Depression. However, its effect on creating respectively car-oriented and flight-oriented suburban sprawl as typologies are predictable. Sizable cities won’t evaporate due to this (though smaller towns may).
– Cities “turned inside out”, industrial land use moving from the waterfront and near-downtown to the suburbs. I’m not sure if I know of any group calling this in advance (e.g. the Garden Cities movement imagined their cities would have their local, near-downtown industrial parks). But a century is a lot of technological development, so for this class of thing, throw out something to the effect of “maybe something similar to retirement sprawl will eventually become much more prevalent, caused by mumble (telework/IT/AI/UBI)”.
– There were also even greater changes to particular industries, e.g. the continued development of refrigeration allowed meat processing to move backward along the supply chain. No more trainloads of livestock into a “meatpacking district” inside cities. For this category, I don’t have anything interesting.
– Urbanization, and in particular the growth of megacities, will continue.
@ Basil Marte
We agree that you need to provide 24×7 service at least every half an hour all day. You need to do regular maintenance on your train tracks. I also agree (though it may not have seemed that way) that cost of buses vs cost of building the crossovers is important to consider.
Where we disagree is there the cost benefit comes out. I weight the long term operating costs of a bus higher, along with the risk that future you will decide to cut costs by not running that service higher. I also think you should be able to build those crossovers cheaper than the US is building them – though here I will admit I’m not sure.
Of course when I consider operative costs of a bus I am assuming driverless busses are in the very distant future. There has been great progress made here though, and they may be closer than I think. (the only people talking about driverless cars have reason to “lie with statistics”, but they are looking good compared to human drivers in their numbers) The limit for driverless cars seems to be the costs of mapping every possible route, but a bus should run on fixed routes and so you don’t need to map anything else so they may be easier to roll out. Only time will tell how this plays out, but I suspect once/if this technology comes out it will change everything.
Some roads remain as busy roads for centuries. However I have also seen in historical documents that some road that is now a dead end used to be very busy. (often because when the freeway went in – cross streets without a bridge, or parallel streets that no longer have traffic). Probably the majority of busy streets stay that way, but a small number no longer are and predicting which will be which in the future is not really possible.
@Henry, given only some of the least well run systems in the world (NYC and Chicago) and the brand new Copenhagen system do 24 hour running I suspect it isn’t really a good idea and that night buses are better.
Copenhagen might well change once their system is old too.
@henrymiller74:
The hard part right now is finding bus drivers, period.
The U.S. custom is for bus drivers to bid on assignments based on seniority. Since union agreements try to give each driver 40 hours of work a week, this is accomplished through low-quality assignments like split shifts (drivers only work revenue service during rush hours, sit in the garage during the daytime between the morning and evening peaks, and drive again in the peak) or graveyard bids.
These dogcrap assignments go to the lowest-seniority drivers.
The Chinese requirements are at least one crossover per 5km and one storage loop per 10km for 80kph lines. For faster lines, the required crossover and storage loop densities are relaxed to one per 8km and one per 15km, respectively. Crossovers are “recommended” at stations near tunnel portals and river crossings.
The crossovers mainly come into use when a line needs to run short due to extreme weather or other misfortunes; the storage loops store extra trains or turn some trains back to boost capacity (eg. during rush hour or after soccer games), and can store broken down trains, ending their disruption to normal service.
There’s only one line in mainland China to build extra crossovers to enable 15-30min service 24/7: Xiongan R1 line, a 200kph airport line.