Los Angeles: Should Cut-and-Cover Be Used More Widely Today?

Reviewing a video from Nandert and talking cut-and-cover.

Next Metro is a blog about public transit expansion, projects, and systems around the world. Make sure you’re subscribed to support the blog and so you never miss a post!

---

Note: You should probably expect more posts on LA going forward, since I will be doing more coverage of North America, and LA along with Toronto and Seattle have the biggest expansion programs on the continent.

Nick Andert has long made excellent overviews of the substantial transit expansion going on in Los Angeles, with amusing quips and very nice animations and maps.

Even better, he’s started doing even more more direct project-based technical advocacy pushing for specific technologies for the Sepulveda line (ok, in this case mostly pushing against the unbelievable monorail options), proposing a new crosstown metro line of his own design, and now promoting cut-and-cover for the northern extension of the K line (formerly the Crenshaw-LAX line) in a new video.

The video is long but definitely worth watching, as Nick goes over the entire alignment of the northern extension of the K line beyond the D (formerly the Purple line) line and to West Hollywood and the B (formerly the Red line) line. Naturally, the big push is to do this project cut-and-cover as opposed to the sort of bog-standard tunnel boring machines and mostly cut-and-cover stations approach that is the default for modern transit construction in North America.

Now, cut-and-cover sort of famously is an old and potentially disruptive process, and while I frequently talk about it being a good idea because my hometown’s (Vancouver) very successful Canada line project used it, it is worth noting that its use did lead to a bunch of lawsuits (which were probably still far cheaper than more expensive tunneling and deeper stations would have been). I think the fact that Nick lives in an area served by the route and is personally pushing for cut-and-cover is very important.

Whenever people talk about the disruptive nature of cut-and-cover, I always like to remind them of a few things.

For one, stations are often built cut-and-cover even when tunnels are not, so some degree of disruption and use of the technique is inevitable.

For two, cut-and-cover is faster. People often talk about how we used to build subways more quickly, and part of that is just because what we were building was not so excessive. Since the work is faster, even if it’s more disruptive, you also should consider that overall disruption is probably similar, just more intense and for a shorter period with cut-and-cover.

Somewhat anecdotally, people tend to prefer “getting it over with”, rather than prolonging the pain (including in transit construction examples), and that’s basically what cut-and-cover does.

A huge part of the push in this case is based on a desire to complete the project faster; this is in part because if the project is less expensive, less time will need to elapse for the tax revenue to fund the project to be generated, but also because cut-and-cover should enable more parallel work and faster work at that.

The proposal made by Nick is great because it’s quite detailed, and at the interchanges even includes information about the depth and rough extents of the existing subway infrastructure for the B & D lines — which Nick suggest going over — which might be challenging, but really would be such an improvement, because given there is roughly enough space to go over, you can only imagine how deep the K line interchange platforms would be if the line went under.

It’s hard to overstate the importance of this project. While right now the K-Line is a bit like a shuttle from the E (formerly Expo line) line south to LAX (via the new connector station I talked about in a post last month), its northernmost section is already underground.

One of the biggest issues with LA Metro as it stands today is that the network is highly radial around downtown LA, despite LA being very weakly-centred. This is made even worse because many of these radial lines are really long. In practice, this means that trips between the ends of the B, D, and E lines to locations anywhere besides downtown LA require using connecting buses or incredible amounts of backtracking (which people hate).

As the image above shows, the K line northern extension starts to create an actual subway grid in central LA west of downtown that should make tons of trips more direct, and also divert passengers away from downtown LA. It will also help reduce the need for multiple transfers — for example, a trip from LAX to Hollywood today would require taking the K line to the E line, the E line downtown to the B line, and then the B line, but in the future could be completed entirely on the K line. Suffice to say, this is a real network-enhancing project that will not only have substantial value in and of itself, but which will make huge parts of the network more valuable for more journeys.

Given how imperative this project is for improving urban rail in LA, I’m really happy to see people pushing for options that get it delivered sooner, and also bring the benefits forward. It stands to reason that future funding for public transport in LA hinges to some extent on current projects completing and having a positive impact as well as more people using and seeing the value of public transit — the K line northern extension has the ability to be the catalyst for all of that.

---

To be clear, I love Nick’s proposal and I would be happy if it were built as suggested, but I think it could be made even better, and should be because if this project goes well it could be the blueprint for all manner of additional subways in the region.

I think it would be wise to look carefully at the Canada line for this project — not only because it is also a modern cut-and-cover subway on the west coast of North America, but also because it has a similar design capacity to the northern part of the K-line (that is, decidedly light metro, as opposed to a full-fat subway).

Probably the biggest learning to take from the Canada line is that some of the proposed stations (especially lesser-used ones that can be added back in due to lower costs — fancy that) could be more minimal. Most stations on the Canada line have a single entrance, which is a pavilion on one side of the street —

Broadway City Hall Station

— while I appreciate Nick’s proposal to have almost every station have entrances on both sides of the road, it seems to me that some more flexibility would be good, and I think that having an entrance on one side and a street redesign is sort of the best of both worlds for a minor station. From a capacity perspective, virtually all of the stations would be fine with a single entrance (and a second emergency-only exit), given that the Canada line has that setup and it alone moves more than 50% of the ridership of the entire LA Metro rail system. These entry pavilions could be developed on top of, and would add another tool to the toolkit.

King Edward Station By Northwest - Own work, CC BY-SA 4.0, https://commons.wikimedia.org/w/index.php?curid=79045659

While some stations might avoid property takes and reduce cost by having stairs on the sidewalk, if you can’t do that, a pavilion is sort of nice, providing better weather protection and a place to wait or meet people. I do think it’s also worth pointing out that a nice benefit of cut-and-cover is that since the stations are shallow, you could build boxes for some stations and infill them if adjacent developments occur.

Nick also goes to great lengths to keep the corners between the streets the K line would run under fairly broad, but I don’t think that’s really necessary. The high-floor “city trains” used on many LA Metro “light rail” lines are derived from trams, which has pluses and minuses, but being able to handle tight curves is a real strength of these trains. Given New York, London, and Toronto have loads of tight curves on their true subway systems, I think it’s worth a deeper analysis to confirm whether trying to keep them wider is even worth it, in part because the time savings are probably small enough that they are eaten up by the unfortunately long headways that Metro often runs. For what it’s worth, the Canada line has a section where it takes a number of tight curves, and while that slows things down — as with the K line northern extension — it’s still going to absolutely obliterate a car in even moderate traffic (which actually has to stop at a corner at least some of the time to wait for a light!).

If LA wants to build more cost effectively, it should probably also nail down just a few modes of rail to serve as many needs as possible. Given that the high floor light rail trains on the A, C, E, and K lines can be made long enough to be similar in capacity to the trains on the B and D lines (especially operated at those annoyingly low frequencies), it’s probably worth doing future subways as “light rail” so that skill can be built constructing that specific type of subway — Edmonton could be a good model with its five car-long high-floor light rail trains and downtown subway, while Frankfurt has such trains with a design that allows a fully-open gangway (letting you walk inside the train from end to end), creating all of the conditions of a nice modern subway.

In such a scenario, if a project justifies something more than a five-car light rail subway, I’d suggest something even higher capacity than the B and D subway lines (8 cars, designed for high-frequency operation) and also with higher speeds — crucial for the long distances in the LA region.

This would create some good segmentation where a “light rail” subway has roughly 115 - 120 metre long platforms, and an “express regional” subway has roughly 200 metre platforms.

Another thing that could enhance this proposal is using a few cut-and-cover adjacent tools — not because they are necessary (albeit some sections of street do seem a bit tight for cut-and-cover and keeping the suggested lanes open), but because they could mitigate disruption and speed construction even more. The Milan method — which is essentially upside-down cut-and-cover where you build the roof and walls of the tunnel and then excavate the tunnel itself afterwards, minimizing disruption — could probably be put to great use. I also think that potentially stacking the tracks vertically could reduce the tunnel footprint — reducing surface disruption and being totally feasible in between the big interchanges (which have crossovers).

Possibly the most remarkable thing mentioned by Nick in the video, which so clearly makes the case for cut-and-cover, is that the entire volume excavated for the cut-and-cover K line proposal would be less than the volume of just the stations in Metro’s proposal. This is largely because the current plan (as with most North American TBM subways) is to have the stations very deep underground, which naturally multiplies by the size of the station footprint. I think Nick could have gone into a bit more detail about this, and I think he may have in a previous video, but his improvements are not just limited to the tunnels, but also substantially shrinking the stations — which might lead to pushback. Modern subway stations in North America so often feature tons of underground rooms, which are obviously expensive to build, sometimes are incredibly excessive (separate break rooms and washrooms for different types of staff), and often ought to either be placed on the platform (which generally has more than enough space since these builds are often grand) or above ground, perhaps even in an adjacent development. While cut-and-cover is great, tackling the station sizing problem — both due to depth and excessive underground space for things that either do not need their own room, or do not need to be underground — is arguably just as important (It’s also seriously worth asking how older subway systems / lines / stations that had far less “back-of-house” space manage to function today. The costs of these huge underground station volumes have gotten so high that you probably have a huge possible solution space within the same or lower budgets).

It is also probably worth highlighting the absurdity that is these giant stations, and then the mismatch with the modest capacity of the K line (which I believe is only designed for three-car trains like LA’s other high-floor light rail lines). I’ve complained a lot historically about how LA builds these lines with far too little capacity, so that if transit ever really takes off in LA, they will suddenly have way undersized trains. But what’s crazy is that many of the new underground light rail stations in LA manage to both still be designed for three-car trains (not big), but also be massive structures. Either build these lines for five-car trains (which could easily fit in the envelope of many of the builds, especially with the “crossovers everywhere!” approach) and still slim the stations down, or drastically cut station sizes and keep the three-car design.

I do actually think I could make the case for three-car trains on lines like this in LA, but only if we were going for the Madrid subway model where we build a ton of relatively low-capacity lines (ironically, Madrid’s “low-capacity” subway lines have much more capacity than these lines in LA, because despite having similar-sized or smaller trains, the service is much more frequent).

You could imagine in LA this might look like a ton of Canada line-type projects forming a very fine grid. Of course some lines would need to be higher capacity (those cutting across larger parts of the region), but most of them wouldn’t need to be!

Currently, LA is both not building enough parallel lines (funnelling riders onto just a couple routes), and is under building them in terms of capacity (for a future LA with much higher transit mode share).

With all of that out of the way though, I am really happy this video, and this thread of advocacy (which builds on solid cut-and-cover advocacy for the Roosevelt Boulevard subway in Philadelphia) exist. Cut-and-cover is used on modern subway projects all around the developed world (and the developing world, much of which has better subway networks than even the biggest North America cities!) and makes a ton of sense here. This is doubly so for LA, which possess pockets of really nice human scale walkable urbanism and also an incredible amount of road space to build on. Plus, since LA is going to naturally have a pretty griddy subway system, the issue of weaving under and over two or three lines simultaneously (which often requires boring or mining) is not going to be present — with strictly north-south and east-west lines, you ideally are just crossing other shallow subways (if rail is even underground at all).

What seems clear to me is that LA needs to push for more cut-and-cover (while also pushing for smaller stations and other cost control-related policy), while also pushing for far more subways (that is, underground rail), including even undergrounding parts of the existing light rail lines (the Flower St. section of the A and E lines in south downtown for example), and completely converting them to subway — along with longer platforms to meet the demand that higher subway frequencies and speeds would bring. LA also clearly needs to get great at building SkyTrain style elevated rail, but that’s a topic for another day (subscribe so you don’t miss it!).