Cut-and-Cover is Underrated

Subways can be built in two ways: cut-and-cover, and bored tunnel. Cut-and-cover means opening up the street top-down, building the system, and roofing it to restore surface traffic; bored tunnel means opening up one portal and digging horizontally, with less surface disturbance. In the last generation or two there has been a shift toward bored tunnel even in places that used to build cut-and-cover, despite the fact that bored tunnel is the more expensive technique in most cases. Regrettably, people don’t seem to even recognize it as a tradeoff, in which they spend more money to avoid surface disruption – some of our sources have told us that avoiding top-down cut-and-cover is an unalloyed good, a kind of modernity. Even more regrettably, this same thinking is common in much of the developing world, where subways tend to be bored.

What are cut-and-cover and bored tunnel?

Cut-and-cover refers to a family of construction techniques all of which involve top-down tunneling. In New York, one of the sources cited on NYCSubway.org refers to the subway as “a covered trench” rather than a real tunnel. The oldest cut-and-cover subways were dug by hand, but in the last 100 years there have been technological innovations to mechanize some of the work as well as to reduce surface disruption, which is considerable and lasts for a few years. These innovations include the cover-and-cut system invented in 1950s Milan (“Milan method”) and the caisson system used to build T-Centralen in Stockholm. The Milan method sinks piles into the street early and builds retaining walls to allow for truly vertical construction, whereas traditional cut-and-cover must be sloped, which requires a wider street than the tunnel, like the Manhattan avenues or Parisian boulevards but not Milan’s Renaissance streets. The caisson method builds a concrete structure and then lowers it into the ground, which facilitates multistory cut-and-cover structures at transfer stations.

Bored tunnel involves digging just one portal, or sometimes a few to speed up work, and then drilling horizontally. This used to be called a tunneling shield, but the shield has been automated to the point that a small crew, only 8-12 people, are required to supervise it nowadays, and now it is called a tunnel-boring machine, or TBM. This method was first invented in London for the construction of the Thames Tunnel, and has been used for all of the London Underground lines since the first two, as London lacks for wide streets for cut-and-cover work. Most American, European, and East Asian cities have switched to this method in the last generation; thus for example New York started to build Second Avenue Subway in the 1970s cut-and-cover, but the program since the 1990s has always been bored.

The typical method used in the world is really a mix – the tunnels are bored, the stations are cut-and-cover. This is because, while the TBM is capable of building tunnels easily, it cannot build stations. Mining or blasting a station is expensive, and many modern examples run up to $500 million or more, not just in high-cost New York but also in otherwise low-cost Rome. This mixed method involves opening up the street at station sites for 1.5-2 years in Paris, intermediate costs, and disruption only at sites that would benefit from the opening of a station.

How much do these techniques cost?

The cost of a mined station starts at $500 million and goes up. But very few cities mine stations – New York and London do, and very rarely other cities do in constrained historic centers like Rome’s. The typical cost of bored tunnel is much less; the lines for which we have seen a breakdown in costs between tunneling and stations, which are a small fraction of our database, have tunneling costs ranging from around $50 million per km to somewhat more than $100 million per km, not counting systems, overheads, or stations. With everything included, this should be viewed as about $200 million per km; the actual median for subways in our database is about $250 million/km, but it includes expensive lines with mined stations, city center tunnels that can’t easily build cut-and-cover stations, and projects that are unusually bad.

Cut-and-cover is generally cheaper. The only cut-and-cover example in our database from Paris, the Line 13 extension to Courtilles, cost 83M€/km, which is around $130 million/km in today’s money; other Paris Métro extensions from the last 15 years are 50-100% more expensive, and the next tranche is even costlier, as Parisian costs are regrettably increasing. Low-cost cities in Southern Europe bore the majority of their subways, but their suburban subway extensions are often a mix of TBMs and cut-and-cover, which is one of many reasons they have low construction costs and Paris does not.

Bear in mind that the superiority of cut-and-cover to bored tunnel depends on the presence of an at least moderately wide straight street for it to go under. London ran out of such streets after it built the Metropolitan line; the District line was, per Wikipedia, three times as expensive, about $110 million/km in today’s money, because it needed to demolish property in Kensington, already then an expensive neighborhood. New York used bored tunnel to cross under rivers and under the hills of Washington Heights, switching to cut-and-cover elsewhere; readers who have gone to the New York Subway Museum will remember the exhibits about the dangerous work of the sandhogs underwater. However, that bored tunnel was no more expensive in turn-of-the-century London than cut-and-cover was in contemporary Paris and New York does not mean these relative costs persist today. Today, on the sort of streets most cities build subways under, cut-and-cover is cheaper, by a factor that appears to be 1.5-2.

The situation in developing countries

In developing countries, I am not aware of any cut-and-cover, which does not mean there isn’t any, just that in the places I’ve looked most closely, namely India and Thailand, the tunnels seem bored. Of note, both India and Thailand build extensive elevated networks, so their subways are to some extent built where elevated construction is infeasible or undesirable. However, to some extent is doing a lot of work here. The Bangkok MRT goes under Rama IV Road, which is about 35 meters wide, and under Asok, which is 30 meters wide. This is comparable to the Sukhumvit, a 35-meter-wide road that hosts the BTS el. Deep-level construction is not necessary on the main roads of Bangkok.

What of other developing-world cities? Bangkok may be unusual, in that it’s a solidly middle-income city, the dominant capital of a middle-income country with comparable GDP per capita to China. What of genuinely poor cities? At least in the bigger ones, wide boulevards for cut-and-cover are not in shortage. Nairobi has vast roads hosting matatu routes. Lagos has such wide main roads that when I crayoned it I proposed that the main radials be elevated, as the under-construction Blue Line is, to avoid having to tunnel underwater from the mainland to Lagos Island. In most cases, short bored segments may be needed, or else short segments that involve the purchase and demolition of private property, as happened in New York when the city carved Seventh Avenue South and Sixth Avenue through the Village.

I suspect the reason this is not done is that planners believe that TBMs are more modern. The physical TBM is an engineering marvel, and looks like advanced technology, even if what it produces is comparable in quality to what cut-and-cover could do when there are wide roads to tunnel under. Planners in the United States have treated it as a given that it’s better to avoid top-down construction. This isn’t even isomorphic mimicry, in which poor countries improperly imitate rich ones; this is proper imitation of a technique whose use in rich countries too is often in error.

Cut-and-cover is underrated

Instead of tunneling wherever possible, I would urge urban subway planners to look to cut-and-cover more. In poor countries, it can be done with the same labor-intensive techniques that produced $40 million/km subways (in today’s money) in New York and Paris. In rich ones, it can be done with more advanced technology to save labor and keep costs under control. This involves more surface disruption, but this disruption can be mitigated by using the Milan method on roads that are wider than those of the center of Milan, and the ultimate benefit is that a lot more subway can be built.

111 comments

  1. michaelrjames

    The physical TBM is an engineering marvel, and looks like advanced technology, even if what it produces is comparable in quality to what cut-and-cover could do when there are wide roads to tunnel under.

    Just a teensy comment. The remarkable marvel about TBMs is how they can tunnel thru strata that look close to impossible, like water-sodden sandy clays or mud etc. The other big attraction to city administrations and planners is how it avoids the rerouting or rebuilding of all the complicated existing sub-surface infrastructure cities have like sewers, water supplies, gas pipes, electric supplies, comms etc. In old and big cities like Paris, NYC etc there is also the complication of many existing tunnels so that anything new really needs to go deeper to avoid them too.

    Also, I’m a bit surprised you haven’t compared costs from tunneling using wide tunnels so as to incorporate stations–because tunneling cost doesn’t scale in simple proportion to diameter (as I am sure I learnt from you). Isn’t this shown by the contentious San Jose BART extension:

    https://www.sanjoseinside.com/2018/04/02/bart-backs-one-tunnel-option-for-downtown-san-jose-extension/
    BART Backs One-Tunnel Option for Downtown San Jose Extension
    By Jennifer Wadsworth, April 2, 2018
    The long-awaited BART extension to San Jose’s urban core will involve one tunnel instead of two—a big win for downtown businesses.

    The double-bore design would feature parallel channels linked by a passenger platform. It would lie closer to the surface than a single-bore tunnel—about 35 feet below ground as opposed to 85 feet—and would require an estimated $70 million more to construct.
    A single-bore tunnel, by contrast, would cost less to build because it would bypass major public utility infrastructure. But it would cost slightly more to operate over the long-term. The VTA puts the price of running a single-tunnel system about 2.8 percent higher over a 30-year period than the two-bore alternative.
    Building a two-tunnel system would not only cause significantly more surface-level disruptions, it risks a political toll as well by potentially undermining public support for a project that’s already behind schedule and over budget.

    • Alon Levy

      Re tunnel diameter, it’s complicated. The actual boring is roughly proportional in cost to diameter. The problem is other stuff. Building subsidence is more difficult to deal with when the diameter is larger, which is why Rome rejected the use of the Barcelona large-diameter TBM method. San Jose, of note, is spending more on the Barcelona method even under a wide street without prior subways, because the Chamber of Commerce complained and chambers of commerce are sovereign or something.

      • michaelrjames

        @Alon: “The actual boring is roughly proportional in cost to diameter.”

        Right, but what I should have said is that the increase in diameter to carry the extra tracks, or two trains vertically is not much more than a side-by-side double-track tunnel, at least with modern TBMs. Nothing approaching 2x. That’s independent of the big savings on mining stations. In fact the Chinese quoted something like 25% extra for a single tunnel big enough to carry two levels of two trains each (ie. 4 tracks), compared to a single two-track tunnel, to traverse Sydney Harbour. (The NSW govt didn’t even consider it because they want to privatise both the WestConnex road tunnels and ultimately the NW Metro so don’t want their operations using shared infrastructure even though it is taxpayer-funded and -built in the first instance. There is also the concept of such a megatunnel providing for future HSR but that is way too much future planning for any Oz entity.)

        Maybe there was the usual nonsense but the San Jose thing was comparing a single large-bore tunnel versus dual single-track tunnels.

    • Richard Mlynarik

      You you, there’s never been anything to indicate that my Aussie schoolboy tourist French is inferior to michaelrjames’, what with the latter never citing any French primary sources.

      On the other hand, I am fluent in Californian — and not just in googling in English for keywords of publicity materials publicized on web sites in California — and can very confidently state that there’s never been any BART capital project ever, at any time in history, nor will there for any time I will see to see, that has not been totally controlled by the concept of cost maximization.

      (This may be one of those subtle linguistic/cultural things nobody could have fully conveyed in your adult continuing education classes in American English. Dommage!)

      Anyhoo … the most basic deal with “tunnels” is that “risk” is the zero-th order cost. “Surface area” (proportional to diameter, ie α concrete lining volume) is, in the best and most professionally engineered case, second-order. “Volume” (α^2, ie spoils transport) is almost in the noise, especially in a barely-urbanised wasteland such as “downtown” San Jose, California, USA.

      Oh, and best way to minimize the dominating cost? REDUCE THE LENGTH OF TUNNEL. But googling “teensy” “thru strata” isn’t going to hit that result. Google on, dude!

      • Gok (@Gok)

        >there’s never been any BART capital project ever, at any time in history, nor will there for any time I will see to see, that has not been totally controlled by the concept of cost maximization.

        Hol up. Surely the Market Street Subway coud have been made more expensive if it were bored instead of cut-and-covered. Or if the Transbay Tube were bored instead of immersed tubed.

    • Tonami Playman

      Didn’t VTA increase the tunnel diameter from 41′(12.5m) to 55′(16.8m) so as to allow island platforms within the tunnel before going back to the initial 12.5m diameter with stacked platforms.

  2. Adam

    Is cut and cover cheaper because everything is shallower? If everything is shallower is there sufficient clearance for all the various utility relocations? Are utility relocations an even bigger and more expensive proposition for a fully cut and cover route? What amount of leverage would this give the many unions that are required to be used to relocate utilities and how would that impact the already fraught and difficult schedule sequencing of utility relocations?

    It’s a terrific idea and I deeply wish if it were used that it would realize the cost and schedule savings it should. but the cynic in me expects if it were attempted, the agency would be forced to cut and cover at least 1.5 x deeper than they want (for specious and excessive regulatory clearances from relocated utilities and any potential cost savings would mostly evaporate from that extra depth; the additional utility relocations would chew up most of the rest of cost savings. And by making utility relocations more extensive, any savings in schedule would be eradicated by that as well.

    And then seven years later, just after the line opens three years late, NEPA will issue a regulation revision that states the faster emergency egress from shallow cut and cover is inferior to the slower evacuation from deep single bore tunneling and no other cut and covers will ever be built. Haha.

    Ultimately the mega conglomerate international firms bidding and building these lines now expect all of them to be extremely high cost and expect that cost inflation for new lines will be at least 15% per year. There is no way with this amount of money on the table that any alternative that profoundly disrupts their wonderful money-printing machine will be allowed to coexist. And since ever more expensive lines keep getting funded costs will continue to escalate at their insane pace. And the mega firms will continue to be extremely bloated and happy and will continue to raise the “rent” of building expensively.

    I am jealous of the short time frame of Paris station construction you mentioned above, I was working between two of the new purple line stations in Los Angeles when both started construction, in 2015 and 2016, 6 years later neither station is complete. And both began construction after three years of utility relocations. Next year, 2022 will mark ten years of street disruption construction at each station, only a couple years more to go after that!

    • Tonami Playman

      @ Adam “Are utility relocations an even bigger and more expensive proposition for a fully cut and cover route?”

      Can the digging not be done around the utilities suspending them with truss trays above the dig while the digging continues below? That way the utilities can be isolated without service disruption and covered back above the tunnels once construction is completed.

    • Alon Levy

      I am slowly learning about the costs of utility relocation and the issues involved therein, but so far the problem looks vastly overstated.

      Of note, people bring up utilities as a reason not to do cut-and-cover even at stations in the United States, even in very young cities like San Jose.

      Parisian metros do not, sadly, take just 2 years to build. The M13 extension to Courtilles took 3, which is unusually short by French standards. But cut-and-cover stations such as for M12 tend to involve digging in the open for 1.5-2 years; overall construction takes longer because construction includes other things than the dig, and there’s real effort made to reduce the dig length to limit disruption to a reasonable minimum. (The French state occasionally does good things, it’s not just banning civil rights organizations and shrugging off a deadly pandemic.)

      • Richard Gadsden

        Yeah, if you want to minimise surface disruption, you dig the hole as quickly as possible, put in the major construction elements (walls, platforms etc) and then cover it over. You still have to do all the cabling, signalling, lay the track, put in all the electrics, tile the walls and floors, install ticket machines, etc, etc, but you can do all of that with workers going in and out through the passenger entrances – and that can take longer than the digging, but the disruption is now confined to a compound at the entrance rather than being a giant hole in the ground disrupting everything around.

      • Henry

        To add on to the sentiment that utility relocation is overstated, but I really don’t think NYC really saved any utility relocation by doing bored tunnels on SAS; they still did a lot of it.

  3. anonymouse

    In Mexico, earlier construction was largely cut and cover and later construction tends toward bored. In Guadalajara, lines 1 and 2 are cut and cover, line 3 is bored through the city center (but elevated elsewhere). In Mexico City, almost all the lines older were built cut and cover except for the southern part of Line 3 (bored due to local topography I believe). Line 12 which is the newest and has entirely bored tunnel (but still has massive disruption due to station construction), and Line 7 older and entirely bored tunnel as well, partly due to local topography and geology, but also it just so happens to run through the posher side of town.

  4. Herbert

    There is another reason cut and cover has fallen out of favor:

    Back in the heyday of cut and cover, trams were largely seen as “not modern” and at the very least networks weren’t expanding. But where a cut and cover subway can be built, a tram can be built for less disruption, less money with the only real “downside” being that to run a good tram you’ll have to permanently take away space from cars – something that was anathema in the 1970s but is seen as de rigeur in the 21st century in more and more places.

    There is of course a constraint of cut and cover you are oddly silent on: Cut and cover has to follow if not roads, then at least sections of non built up space or if it goes through built up space, space where tearing down the building (and paying whatever the laws say must be paid to the owners) is cheaper than using bored tunnels instead. In many cases subways are “sold” to the public as opposed to trams for two reasons: They can carry more people than trams, and they can go wherever independent of surface space availability. A cut and cover subway is much more constrained in where it can go. And the “But when it’s done, there’s more surface space” argument falls flat as during construction that space is consumed. And why should what works temporarily not work permanently? Meaning turning over that road space to non-car uses…

    That said, the German cities that still build new tunnels for urban rail still occasionally make use of cut and cover where feasible. IIRC the U3 southwest extension in Nuremberg (from “Gustav Adolph Straße” to “Gebersdorf”) uses/has used cut and cover on some sections – but not all due to the above mentioned constraints…

    • Eric2

      There is a huge difference between metro and trams. Metro can be ~twice as long (stations are not limited to a city block) and ~twice the frequency (no grade crossings where the traffic light is red half the time). They are also faster due to the lack of surface obstacles and because stations are generally less frequent, which lets them serve longer range trips. They also have the option of driverless operation.

      Now you may be thinking of German conditions where every high demand corridor already has a metro and where there are no megacities to begin with, so the extra capacity of metros is unwarranted for any future lines. But in most of the rest of the world, metros are a necessity not a luxury.

      • Herbert

        The best metro frequency I know of in Germany is on the parts of U2 in Nuremberg where it overlaps with U3. There the nominal headway is 100 seconds (obviously there is a fudge factor – sometimes two trains may be 90 seconds apart and sometimes 110)

        Now I never did stand at that stop with a stopwatch to time it, but at Walpurgisstraße, eyeballed as the busiest Dresden Tram stop with two tracks only (others are + shaped or L-shaped) five lines overlap https://www.dvb.de/-/media/files/liniennetz/dvb_lnp_city_250121_pdf.pdf (see network plan) and from the https://www.dvb.de/de-de/fahrplan/haltestellenauskunft?stopid=33000003&date=26.02.2021&time=11%3A59&time_submit=%5Bobject+Object%5D online departure information we get ~ 1 tram per minute per direction at peak. Of course at such short time intervals, we need more than eyeballing, but I think it is not entirely unreasonable to assume there is a similar capacity for short headways on both subways and trams albeit perhaps with some benefits to a ROW that cannot be intruded upon on the one hand and some benefits to a system that does not usually need signals on the other hand.

        Now as for the possible maximum length of vehicles, first of all, it is the exception, not the rule that cities are built neatly in “blocks” – and even more the exception that those are of any kind of standard size. At any rate, the true limiting factor for the length of any rail vehicle is platform length. If you, like Nuremberg U-Bahn only build platforms for four car trains (and even run a significant share of service with two car trains) you cannot run six car trains. Many tram cities are retrofitting existing stops to allow longer trams as demand has risen. And I think such a retrofit is a lot cheaper on the surface than underground.

        Are there corridors where you have to have a metro? Of course there are.

        But there are far more corridors where a metro is not really needed but a bus cannot possibly cope with the demand. And that is where trams come in. Certainly at a certain range of trips, the tram is “faster” (door to door that is) because one does not have to go down (or up) to get to the train and the reverse once getting to the destination.

        • Herbert

          When travel becomes possible again, I will venture to go to the aforementioned Dresden tram stop stopwatch in hand and report the results on this blog’s comment section….

        • RossB

          “But there are far more corridors where a metro is not really needed but a bus cannot possibly cope with the demand. ”

          I really doubt that. Just about every Metro I’ve seen has been justified based on speed, not capacity. Once you spend the money on the infrastructure to make it fast, it often makes sense to run rail (especially if it is a closed system). In North America, the trams are not built because of capacity. They are built because they are cute. There are very few places where buses are at capacity. The one place where it would really make sense is in Vancouver (UBC Line) but that will be replaced by a subway (which is the right choice — too many cross streets).

          You could certainly make the case that “surface is underrated”, but it is just one tool in the toolbox. A lot of subway systems are a mix — they have deep bore, cut and cover, surface and elevated rail. It really depends on the corridor. Seattle’s light rail line has all types. Parts of the original bus tunnel was cut and cover. This is now the core of the subway line. All the rest of the tunneling was deep bore, and it rarely follows the street grid. Much of it is elevated. The section that Henry referenced below is on the surface. It is worth noting that the area does not have major cross streets. By running on the surface, the train has to have signal priority, and it doesn’t always work smoothly. The train does run a bit slower than it could, but if the stop spacing wasn’t so ridiculously large, no one would notice. The headways are limited to every six minutes because the city won’t agree to give the train any more priority, despite it running on the main thoroughfare. They didn’t actually “take” general purpose lanes, but instead widened the street. These were all trade-offs. I think they made the right choice, but these are trade-offs. (By the way, I think the agency is dysfunctional, so for me to say “good job” is saying something).

          It really gets down to the particulars. If you have a long flat corridor with very few cross streets, then surface is likely the best choice. If you have lots of cross streets, then elevated or cut and cover is the way to go (depending on surrounding buildings, width of the street, or other factors). If you need to zig-zag around obstacles, and otherwise make big turns, then deep bore is the right choice.

          Politics, not cost, drives a lot of it. That, and just ignorant tradition. This is what Alon is getting at. Cut and cover is rejected because politicians are afraid of the disruption. Elevated is rejected because people don’t want trains next to their apartment, or they assume they will be noisy. People make comparisons to Chicago, and oddly enough, not in a good way. Instead of “Hey, it works great for Chicago — a wonderful city”, they say things like “We don’t want all that noise and grime, like Chicago” (there is probably a tiny bit of racism in there, but whatever). Surface is rejected because so many American cities have created pointless streetcar lines. They assume they will be slow, when it really depends on the corridor — many are just as fast as if they were underground. If you actually look at the time it takes a rider to get from one place to another, it is actually faster (it takes a minute to go down the stairs to the platform). BRT is rejected for the same reason (to many mediocre bus lines called “BRT”).

          The Langley extension of SkyTrain is a great example. BRT is probably the best choice. My guess is you will never need the additional capacity or rail, and you can run express buses (during rush hour) which would save riders time over any alternative. But if you really need the capacity, then surface rail is the right choice. Neither will happen, and instead they will spend a fortune on extending SkyTrain. Oops.

          • Herbert

            Speed of an S-Bahn will always be better than an U-Bahn. But except for the million+ cities, S-Bahn systems rarely serve intra-city travel purposes (I say having ridden a single S-Bahn stop today both ways instead of the bus) because most trips are too short for them. A tram needn’t be much slower than a subway ceteris paribus. And of course door to door a tram can be faster due to the aspects I mentioned earlier.

            An at-grade fully grade separated system (I know this sounds self-contradictory; just think of a fenced off ROW) is of course maximally intrusive on the urban fabric while being the least costly option from the sole POV of the cost of building the rail line (excluding stuff like crossing bridges or having to re-route buses and whatnot)

            When a subway is built badly, it’ll either leave many people with such a long walk to the subway stop that their total trip becomes even less attractive than the previous bus trip was. Or one ends up running buses anyway which kind of defeats the purpose of building the subway to begin with. A tram can easily railstitute a bus-route without problems. When subways are built to railstitute three or four bus lines (due to cost) or – worse yet – two or three tram lines, there come tradeoffs that are often undesirable… Just compare an old tram map of Nuremberg with what it has now….

          • Sascha Claus

            But except for the million+ cities, S-Bahn systems rarely serve intra-city travel purposes […] because most trips are too short for them.

            Do Stammstreckentunnel (trunk tunnels) like in Munich count as U-Bahn? With frequent service (formed by a lot of routes) and and short station spacing, they can be very useful for intra-city journeys.

            Or one ends up running buses anyway which kind of defeats the purpose of building the subway to begin with.

            Unless there is enough traffic to warrant both a fast, coarse-meshed subway network and a not-so-fast, fine-meshed surface option.
            I’d like to see a map of all tramways that were replaced by a subway and soon after supplanted by a bus every 10 or even 5 minutes …

          • RossB

            “Speed of an S-Bahn will always be better than an U-Bahn.”

            Yeah, but I’m focused on U-Bahn (and I think you are too).

            “A tram needn’t be much slower than a subway ceteris paribus. ”

            Yeah, but you can say the same thing about a bus. It gets more complicated than this, but here is a rough breakdown:

            1) Fast enough on the surface. Don’t need the capacity: Bus
            2) Fast enough on the surface. Need the extra capacity: Tram
            3) Not fast enough on the surface. Don’t need the capacity: BRT
            4) Not fast enough on the surface. Need the extra capacity: Subway

            I’m saying that second category is tiny. In North America, there is only one place I can think of that falls into that category, and it wants the speed of a subway (which I think is the right choice). For example, in Seattle, which is spending over $50 billion on mass transit (on top of what they have built before) there is not a single place where a tram makes sense. There are places where BRT makes sense (everything from bus lanes to tunnels). There are places where a subway would save a huge amount of time. But there are no places where a tram actually makes sense. This is likely true throughout the developed world. My guess is in most of the world where it would make sense, almost all of those places already have trams, or they have a robust subway and bus system. In that case, a tram definitely makes sense, but they weren’t seriously considering anything else (I’m thinking Paris). The only place where I could see lots of (newly built) trams as the main form of public transportation is maybe in parts of the developing world. They have grown very fast, very quickly, and don’t have the money to build a huge new subway system, and the buses are very crowded.

            Of course there can be a mix, and often a mix makes the most sense. Seattle already has light rail running on the surface, and should probably have more. But what they really need, more than anything, is more of that first and third category (better bus service).

      • CJ

        The length of a city block is not really a constraint on station size: if you’re in a world where dedicating former road space to a tram is feasible, you’re also in a world where closing an intersection is feasible. Voilà, a two-block station.

        • Eric2

          I know of tons of places in the world where trams have separate lanes (in some cases I know it was taken from cars in recent decades), but I don’t know of a single place where an intersection was closed to allow for longer platforms. Can you name one?

          • CJ

            Sunnyside and Banff Trail in Calgary; Health Sciences and MacEwan in Edmonton. Without closing the cross street, the stations would either have needed to be shorter or placed in worse locations.

            This happens all the time at non-station locations, too — just look at 36 St NE for Calgary or 111 St for Edmonton.

            It’s not that much of an imposition to make the occasional cross street not a through street, so that you might (gasp) have to detour a block, just like routinely happens for other reasons.

    • Alon Levy

      Three things.

      1. The situation you’re describing regarding subways and trams is far too specific. It’s common in Germany, but the same is seen in countries where nobody thinks a tram is an alternative to rapid transit. Italian and Spanish cities build trams but not as big-city radials; nonetheless, they use cut-and-cover less than they probably should. Singapore has wide roads and no trams nor interest in trams; it bores tunnels and even its cut-and-cover stations are off-street, with demolitions instead of median construction. The US has light rail, but it almost never takes space away from cars – it’s built in freeway medians and disused rail rights-of-way and there are planned Stadtbahn tunnels for it in Dallas and Portland, neither cut-and-cover.

      2. Bored tunnels let subways run independently of the street network, but that’s part of the general fascination with TBMs that leads to their overuse. In reality, in London the early Tube lines ran mostly under streets anyway to avoid getting property easements under private buildings. The same is true of much of the network in Stockholm or Moscow. From time to time one sees a horrifically squiggly TBMed subway, like the Copenhagen Metro or what’s planned for Graz, but I don’t know to what extent it’s truly necessary, vs. just going nuts with tunneling.

      3. Even in situation in which there is no alternative to a subway because the project is an extension of an existing line, running under wide streets, such as Second Avenue Subway or U5 here, it is bored. The plans for phase 4 of SAS even has trains running underneath rather than through a cut-and-cover station box that was built for SAS in the 1960s under Chrystie Street, which is wide and has a park to its east rather than buildings.

      • Henry

        IIRC the SAS Phase 3 Grand St issue was that the station box is wide, but not wide enough now that ADA and including elevators is a thing, and there are a bunch of crickety old tenements on one side of the station that would need underpinning to expand the width. And one can’t simply move the station box, because where it was supposed to go does actually have trains running through it, and one of the supposed benefits of deep Chrystie was less disruption to 6th Avenue-bridge service.

        If it was like Lexington-63rd with trackways behind a false wall, that would be one thing, but AFAIK those don’t exist.

        • adirondacker12800

          The old tenements were there when they built the station. Wikipedia says the station was designed to be converted to two islands when the Second Avenue arrives. With pictures of the existing Second Avenue trackways. With lots of fluff about reducing community impact. Reduce community impact enough it gets too expensive and never gets built.

          • Henry

            Designed to be converted to two islands in the ’50s, before the passage of ADA required elevators.
            The MTA can’t just ignore civil rights law anymore, unless they want to keep losing multi-billion dollar lawsuits.

          • adirondacker12800

            They manage to make other existing platforms compliant.

          • Henry

            And how many of those are also getting turned into cross-platform transfers at the same time? The MTA simply concluded that it was not possible to do both in a satisfactory way, not without significant impact.

            From the FEIS

            An important part of the Second Avenue Subway involves connecting the existing BD service
            at Grand Street to the new Second Avenue Subway service via some type of transfer. Because of
            the narrowness of the existing Grand Street Station platforms, the existing station would have to
            be reconstructed in order to serve the larger volume of passengers that would use the station
            once the Second Avenue Subway service is operational. Each of the three options originally
            considered for the area south of Houston Street would modify the existing Grand Street Station
            differently.

            As a result of information gained through the analyses included in the SDEIS, the Shallow
            Chrystie Option is no longer under consideration, as it would have resulted in more significant
            adverse impacts during construction than the other two options. As is described throughout the
            SDEIS, the Shallow Chrystie Option would have: 1) resulted in more displacement of residential
            and commercial uses, 2) required more underpinning of adjacent properties, 3) caused more
            encroachment into and greater impacts to Sara D. Roosevelt Park, 4) generated more dust and
            noise because cut-and-cover construction would have been required along a longer segment of
            the proposed alignment, and 5) potentially affected possible burial remains at five former
            cemeteries. Although no longer under consideration, the Shallow Chrystie Option was
            nevertheless discussed throughout the SDEIS for comparative purposes.

          • adirondacker12800

            How many of them already had cross platform transfers when they added elevators? Put up with a few years of construction and actually get a station or be blissfully unperturbed because it never gets built.

          • Henry

            I fail to see how that’s relevant, since if a cross platform transfer already exists you’re not physically expanding the station. It’s also not faster to build, and not necessarily cheaper; one of the reasons shallow Chrystie was rejected was because it would also include closure of the F in the area.

            They actually also studied (and rejected) a station alongside at Forsyth St rather than Deep Chrystie, because they projected less ridership due to a longer transfer.

            The Second Avenue Subway’s second phase costs almost $7B and most of the tunnels already exist. There’s a lot more places to fix problems before we decide that a not-very-deep tunnel is the bane of our existence. The MTA couldn’t put together a ham sandwich at reasonable cost if it tried.

          • adirondacker12800

            Because they knew what dimensions an express station with two island platforms should be when they built the IND in the 30s? That they haven’t had to alter to add elevators?
            The F train isn’t anywhere near Chrystie and Grand. Unless they are worrying about Chrystie/Second and Houston. Which isn’t Grand St station on today’s B and D it’s the Second Avenue Station on the F. Which was built to have four tracks of Second Avenue trains go over it, not under it. Wily planners from 100 years ago. Plotting to put the second set of trains OVER the first set. Like they did out in Brooklyn too. And Queens. So that the Second Ave, trains are one level down and cheaper to build instead of three levels down and more expensive.

  5. Tonami Playman

    Alon, In regards to your Lagos metro crayon. “when I crayoned it I proposed that the main radials be elevated, as the under-construction Blue Line is, to avoid having to tunnel underwater from the mainland to Lagos Island.”

    Would crossing the lagoon via a bridge not require a roughly 500m – 600m ramp to transition from elevated to underground on the Island and how does this affect connecting to other proposed tunnels?

    With the land constraint around the central station, would it not be better to locate the ramps on the mainland right before the 500m water crossing and then making the crossing with an immersed tube tunnel?

      • Tonami Playman

        So similar to the multitude of criss crossing elevated tracks around Osaka station.

        • Eric2

          Yes. It’s a big ugly, but when you’re a poor city with severe water table issues, effective transport is a higher priority.

          • Tonami Playman

            Yeah, I hear Bangkok residents initially complained about the aesthetics of the BTS Skytrain concrete viaducts, but have now come to appreciate the shade it provides from the hot tropical sun ( Not only pedestrians, but also car and scooter drivers always try to drive along the shadow of the els). I’m certain Lagos residents would share a similar appreciation.

    • Alon Levy

      You’re gaining around 4 meters naturally on Lagos Island, so the ramp is shorter than 500 meters.

      I’m more worried about the water table in such a flat area, though…

  6. BindingExport

    Within the EU (at least according to the interpretation to some German plan approval authorities) any sort of lowering the groundwater table during the construction of tunnels is practically prohibited or only allowed on a small scale with extensive monitoring. The Regierungspräsidium Darmstadt f.e. is quite open that it won’t allow any tunnel construction within Frankfurt using c&c due to its experience with the last two tunnel projects. Any city with a high water table practically can’t built cut and cover within in the EU due to a ruling by the European Court of Justice in 2015 (Case C‑461/13).

    • Herbert

      Well that’s a shame…

      If they weren’t digging for public transit but instead for lignite, they could of course do whatever they want with the water table….

  7. Transportation Justice CNY

    This is all good information, thank you Alon.

    I’d only add that a shallower station is just better transit than a deeper station. Each additional flight of stairs or elevator ride makes each trip longer and less pleasant. And that cost lasts for decades, unlike a few months of inconvenience to motorists.

  8. Joseph Shupac

    An additional benefit to cut-and-cover might be that you could build a surface LRT route (and other road improvements too) simultaneously without too much extra cost. Having the LRT above might also mean you wouldn’t need to build as many subway stations — the subway would be more express-ish. Of course, all this would be overkill except on streets with extremely high transit demand… or unless cargo trams become a thing…

    • Tonami Playman

      You ramp down and dig a little deeper as you approach the intersection point and ramp back up past it to shallow sub-surface. Lots of these in Paris and some in Tokyo. One example is where the cut and cover Hibiya line dives under the Ginza line at Toranomon station along Sakura dori Ave and again under both the Marunouchi and Ginza lines at Ginza station along Harumi dori Ave.

      But after the first two line intersection, any further intersections have to be bored at lower depths.

    • yuuka

      Treat it like a utility and underpin it, then hope your operations side of things are happy enough with the instrumentation proposed to monitor the tunnel and don’t make any noise.

      But at some point, you’d be so deep that it makes better sense to use a TBM. Not that the typical medium-sized American city is going to have more than two or three subway lines anyway, so they won’t reach that point.

      • Herbert

        Two or three lines cannot hope to cover the “typical mid sized American city”. A dozen light rail lines, on the other hand…

        • yuuka

          Then I’d think this entire discussion is irrelevant unless you’re packing said light rail lines into two or more tunnels – or “subway lines” for the purposes of discussion here.

          • Herbert

            Well I keep arguing that while, yes, there are usage cases where you *need* a Metro, there are far more usage cases for Light Rail….

        • Henry

          Where possible, line transfers should be spread out to provide more coverage and spread the load. So even with a 12-line network I don’t think you’d wind up with many situations with more than two or three transfers at a station, assuming you were following those principles.

          Shanghai has a crazy amount of subway lines and exactly one location where four lines intersect. There are a handful of stations with three lines (Lines 3/4 share track, so I’d count that as one line)

          • Eric2

            Shanghai has low job centralization.

            Times Square, NYC, is a transfer between 5 to 8 lines (depending how you count) and none of them are really avoidable.

            Bank-Monument and Kings Cross-St Pancras in London have 5 lines each.

            Chateles-Les Halles has 7 lines including RER.

            Yes it is good to spread the transfers between stops, but CBDs tend to be extremely concentrated and there is a limit to how many stops you can put on one line there, and thus, a limit to how much the transfers can be spread between stops.

          • adirondacker12800

            Times Square, NYC, is a transfer between 5 to 8 lines (depending how you count) and none of them are really avoidable.
            There is life outside of Midtown. Many of transfers are avoidable, you aren’t trying hard enough. There are 11 lines along with the shuttle during rush hour service. 14 tracks of through service and the shuttle.

      • Joseph Shupac

        Alon, sorry to harp on this, but wouldn’t the ideal situation to cut-and-cover a railway tunnel while simultaneously building an LRT at surface? That way you get the cost-savings of cut-and-cover AND the cost-savings of only having to rip up the road once! Has this ever been done, to your knowledge?

        • Alon Levy

          I don’t think so? The subways I’m familiar with were often built as replacements for streetcars. In fact in Berlin the reason U2 is so squiggly through city center is that the straight route, Leipziger Strasse, had a private streetcar whose owner refused to give permission for the construction of a subway underneath.

        • Henry

          I feel like if you were to go to all that trouble, you might as well spend the money to underground the LRT as well.

          The Market Street Subway in SF is light rail on top, BART on bottom.

          • Joseph Shupac

            Interesting. Was that cut-and-cover built in SF? I wonder how the costs and benefits would compare between: 1. a cut-an-cover rail tunnel, no LRT; 2. a cut-and-cover rail and LRT tunnel; and 3. a cut and cover rail tunnel, with LRT built simultaneously at surface level.

          • Henry

            The Market St Subway is cut-and-cover, yes.
            There is actually a streetcar on top as well, but that came after, and that is more of a heritage streetcar. Although the tracks were in the ground already from past use. (Market Street used to have four tracks of streetcar service; two of them were put underground in the Market St subway.)

        • michaelrjames

          The Hong Kong metro was done cut-and-cover, at least the Kowloon part up Nathan road was, not quite sure about the Island Line. Anyway they certainly retained the famous double-decker trams and its tramway which follows the exact same route as the Island Line Metro below. There were serious suggestions about a decade ago of closing down the tram but it remains popular on that route and it is now run by Veolia rather than MTRC. In such a dense business zone the tram has advantages of hop-on-hop-off so can be more efficient point-to-point than the Metro with its limited stops. The tram gets >200,000 pax per day.

          • Joseph Shupac

            Okay, but what about building a new tram line at surface at the same time as cut-and-covering a rail tunnel? It would be a sort of 2-for-the-price-of-1, right?

          • yuuka

            Island Line is deep bored in the old London Tube style, which means inconvenient access too.

            If they built it cut and cover, the trams would probably have to go.

          • michaelrjames

            @yuuka

            Right. That makes sense.
            In fact I may have it wrong about the Nathan road section in Kowloon. I saw it under construction during my first visit to HK in late 1979 and saw massive steel plates across the road, ie. temporary structure for the road, but now I wonder if it was just for the stations and the line itself might have been deep-bored?

          • Sascha Claus

            Okay, but what about building a new tram line at surface at the same time as cut-and-covering a rail tunnel? It would be a sort of 2-for-the-price-of-1, right?

            Given the price difference between surface and subway, more like 1.1-for-price-of-1 or 1.2-for-the-price-of-one.
            You only need a case where it is known beforehand that you still need the surface tramway / LRT afterwards, *and* where you can convince politics about that! Given that movers and shakers here are overeager to dismantle tramways as soon as a roughly parallel S-Bahn or even Regionalbahn nearby opens, no matter how useless it is as replacement for the tram. It’d be close to impossible to convince such people that somebody wouldn’t take the brand new subway, let alone that they would be too many for a bus.

          • michaelrjames

            @Sascha Claus: “Given that movers and shakers here are overeager to dismantle tramways as soon as a roughly parallel S-Bahn or even Regionalbahn nearby opens, no matter how useless it is as replacement for the tram.”

            Apparently Matt Yglesias in his latest tome, One Billion Americans: The Case for Thinking Bigger, to better compete with China, he recommends building S-Bahn in most American cities! Clearly he’s been reading AL who missed their own chance to write a bestseller (instead of blogs for a bunch of antifa freeloader losers). Who knew where the secret to MAGA lay … (AL, that’s who !).

          • Joseph Shupac

            Interesting. If the politics could be overcome though, then 1.2-for-the-price-of-one, plus the cost savings of going cut and cover instead of bored tunnel (which Alone estimated at 1.5-2 x), adds up to a pretty significant amount of savings.

          • Henry

            Now I could be wrong, but Nathan Road is probably cut and cover, because I don’t know how you’d reasonably manage the whole dual cross-platform interchange followed up by a terminal. Sounds very finicky with TBMs.

          • michaelrjames

            @Henry

            Ever since I walked over the giant steel plates covering Nathan road all those decades ago I have been convinced it was classic cut-and-cover, however I became less sure. It may well have been just the work on the stations. My hesitation was provoked when yuuka said that the Island Line was deep-bored; however that line was built in the years after the first line(s) (along Nathan road) and in fact used the “New Austrian Tunnelling Method” (NATM) which uses a TBM and sprayed concrete (the precursor to using preformed concrete panels). Wiki is no help but I eventually found something:
            https://www.checkerboardhill.com/2012/03/underground-iron-building-the-hong-kong-mtr/
            The Nathan road section was cut-and-cover and the harbour was traversed using immersed tube which was used for all harbour crossings including the most recent Airport Express and Tung Chung lines at the western end of the harbour. The vid shows that sometimes balconies along Nathan road were removed to allow access for machinery. It shows the placement of the steel decking to allow the road to keep functioning above the works.

            Incidentally, even for this first Metro the MTR was involved in building apartments for 50,000 and 2mft2 of commercial space on “their” land.

          • Henry

            I’m not saying you *couldn’t* do it.

            It’s a shame Hong Kong seems to have forgotten how to build these; I suspect Ho Man Tin, Hung Hom, and Admiralty are going to be crapshoots (at least, in the case of Admiralty, more so than it already is). The South Island Line might’ve had one at Wan Chai, but that was dropped on cost grounds.

          • michaelrjames

            @Henry

            Not sure about that (if understanding you). It seems to me it might be the opposite. That part of HK is perhaps the most crowded bit of urban structure in the world. It seems to be certain that deep-bore TBM is much simpler especially as the stations have very complex pedestrian connections to the street and buildings and malls all around. One of the most expensive issues is to have the South Island line share an island-platform with the existing Island line so as to facilitate efficient transfer of pax between them (& to the two other cross-harbour lines at Admiralty; note Wan Chai is not an interchange station). To manage that you’d have to be doing deep station mining no matter which approach. Further, the South Island line had to go thru the mountain so was always going to be TBM.

            Also, in your negative comments I think you are confusing any construction issues with final outcome. I’d have confidence that the outcome will be excellent (if never satisfying monsieur/docteur Levy).

          • Henry

            Well, for starters, the South Island Line *doesn’t* share an island platform with the Island Line. Not sure where you’re getting that from. The station cavern for South Island Line (and the future Sha Tin to Central Link) is not very close to the existing cross platform interchange, but instead at a much deeper level in a new cavern with the Sha Tin to Central Link. (And those lines don’t have cross platform interchange either within their own new station cavern.)

            It’s a bit silly to be shoving everything into Admiralty, when it already had documented issues dealing with the existing Tseun Wan/Island Line cross-platform flows. And there is also congestion at the new escalators with just the South Island Line operations, so it’s going to be a shitshow with yet more passengers trying to squeeze onto the Island Line from another pair of cross-harbor tracks. But of course additional interchanges at Wan Chai, or bringing trains into Central South where everyone is actually going, apparently cost too much money for a consistently profitable rail operator, and so now we move the congestion problem of “everyone is trying to go to Central” south from Kowloon Tong and Mong Kok, to the already congested hub at Admiralty.

            And the problems are already showing. Already there are reports of increased congestion and delays because moving from 12 to 9 cars on the East Rail Line is tanking capacity and confusing passengers, and the promised frequency bonuses from shortening trains have not materialized. (Never mind that you could theoretically just have 12-car trains running at the increased frequencies as well.)

          • michaelrjames

            @Henry

            You’re suggesting Wiki and its sources has Admiralty wrong? The very station interchange symbol (lozenge-ovoid) is reserved for direct platform interchanges; I didn’t say the two other cross-harbour lines were the same and indeed the map shows the correct interconnection (correspondance; ie. two circles with connecting corridor etc).
            The problems you describe at Admiralty are actually the reasons why this connection/terminus of the South Island Line (and future Admiralty to Hung Hom extension) was chosen, and of course why the works are so expensive. It’s the nature of interchange stations and obviously even more when multiple lines converge, but equally the final benefit is so much greater, in network terms. I’m sure those pedestrian movements/congestions are well planned for in the new station complex.

            Here’s the relevant bit from Wiki (its original source is a dead link), though sometimes this can be out of date (however this does seem accurate; the abandoned trying to make the double-island platform for both lines but retained it for the Island Line):

            At Admiralty, a new island platform was built under Harcourt Garden.[21] Transfer passages connect the new station area with the older Tsuen Wan line and Island line platforms, as well as the future North South Corridor platforms.
            At Wong Chuk Hang, the platform structure was planned to be a double island platform with three tracks (like Choi Hung station). South Island line (West) trains would use the centre track while South Island line trains would use those on each side, allowing for convenient cross-platform interchanges. However, according to the final plan, any future South Island line (West) platforms were to be built above the existing platforms.

          • Henry

            When that Wikipedia paragraph says “transfer passages” what that really means is “escalator shafts and lifts”.

            Here’s the MTR’s station map for Admiralty: http://www.mtr.com.hk/en/customer/images/community/art_architecture/architecture_adm_map.jpg

            And here’s a video of an MTR employee demonstrating how to get from the existing Admiralty to the new deep level cavern. It even has English captioning. https://www.youtube.com/watch?v=uvZri4HtMuA

            The Tsuen Wan and Island Line cross platform interchanges already existed from the original system. The new extensions and lines feature no cross platform interchanges except at Hung Hom, and certainly no more dual interchange setups. Though this is less terrible than Nam Cheong, which has a cross platform interchange but in a direction that very few people care to use, and it’s only one pair of directions with a transfer.

            I’m not saying that there should be no transfer at Admiralty at all, but putting all the eggs in the Admiralty basket is penny wise and pound foolish. Considering that some alternatives at least had two possible transfer stations (both Admiralty and Wan Chai for the SIL, both Admiralty and Central South for the SCL)

          • michaelrjames

            @Henry

            OK, I apologise. The Island Line is the deepest of the four lines with the future Hung Hom connection being directly above it. Depending where you are transferring to you could be up for a 4-level trek! In feeble defense I will say the Wiki is misleading/unclear in that it implied they only abandoned part of the original cross-platform concept whereas they totally abandoned it all. Seems like it was just too complex/expensive to achieve, though they still took over Harcourt Gardens at ground level (perhaps it was the TBM portal?). Still, of the four lines, the Island Line will serve the fewest pax so … Also, I remain unconvinced that putting the Island Line terminus at Wan Chai would be better as it would force some pax into taking the Island Line just to transfer to those other lines, thus unnecessarily loading the very busy line & station; now they just need to take some escalators up a few levels! OTOH I would have wished for the Island Line to continue the 900m west to Central (and why not the extra few hundred metres to the HK (IFC/Airport Express) station; the Island Line is deep below all other infrastructure ..
            The current Island Line map on the Wiki page still has the old incorrect arrangement:

          • michaelrjames

            Oops. Posted the wrong link:
            [The current Island Line map on the Wiki page still has the old incorrect arrangement:]

          • Henry

            So I think your confusion lies with the fact that there is both an Island Line, and a South Island Line, which are two distinct lines built at different times with completely incompatible rolling stock. There are also plans for the development of a North Island Line, a third distinct project, except it’s not actually a line in and of itself, but a plan to drive both the Tung Chung Line and Tsueng Kwan O Lines to meet in an alignment parallel to and north of the existing Island Line. (The very confusing naming is probably a British legacy.)

            To be quite clear about what I’m talking about, when I talk about a interchange at Wan Chai, here is a Wikipedia diagram of the plans for these lines circa 2005, depicting Central South, Admiralty, and Wan Chai. These were earlier plans to have a dual transfer system similar to the other ones to spread the passenger loading. Instead today we are just going to get everyone slamming the Island Line at Admiralty, which was congested enough as it is.

          • michaelrjames

            @Henry

            No, I wasn’t confused by those other lines or plans. There really was the intention to build island-platforms to facilitate easy transfer to the Island Line. That would have obviated most of your criticisms of using Admiralty ultimately for 4 lines. Still not clear why they abandoned it except that, as usual for such projects, it was behind schedule and the simplification both saved money and lots of time. Also the old Island and Tsuen Wan lines/platforms may not be deep enough to avoid structural issues with buildings above of any widening to accommodate the extra platform + line.
            Incidentally that figure you linked also shows the same ovoid interchange station at Admiralty, however it is probably not an official MTRC map where that symbol has the specific meaning. That concept–of Wan Chai + Admiralty–is close to mine of the 900m extension to Central, which would mean most SI Line pax would not be transferring at all but heading straight to the exits, simultaneously diluting the load the Island Line carries. These three are probably among the busiest stations on the entire network. In fact you’ll notice that that old map has the future cross-harbour extension to Hung Hom of the Shatin line (still labelled KCR on the old map) terminating at Central (via a stop at Admiralty) while it is now planned to terminate at Admiralty. That actually makes more sense than my suggestion because it would be taking far more pax than the South Island Line.

          • yuuka

            @Alon the distance between City Hall and Raffles Place is much greater than Prince Edward-Mong Kok-Yau Ma Tei last I checked. Also, I seem to recall the stretch between City Hall and Raffles Place also done cut and cover, but I digress…

  9. Onux

    With so few data points (one example in your database in the whole world) are you really sure that cut and cover is cheaper for tunneling proper? I grant cut and cover stations being cheaper than mined is indisputable. As I mentioned in a previous post, the costs of digging from the surface (shoring, dewatering, vertical circulation with lifts/cranes/etc.) grow exponentially as you go deeper. The cost of cut and cover under an existing line (i.e. shoring it up) is expensive, similar to the difficulty of a mined station even if from the surface, and the costs apply regardless if the the infrastructure is rail, road or water tunnel. This suggests cut and cover is only cheap when it can be very shallow with clear underground ROW, like the Paris Metro with its essentially mezzanine-less stations.

    The example of the District line seems to show that even shallow cut and cover can be expensive if it requires takings. Conversely, the metros in Barcelona and Madrid, lauded for low costs, are almost entirely bored. The Spanish method got around the mined station cost with oversize TBMs and in-tunnel stations, but if TBMs were so expensive at moving a cubic meter of dirt compared to surface digging, then the extra cost of larger TBMs should have offset or overwhelmed station savings.

    I believe you are underestimating utility relocation, even if other overestimate it. When London, NY, and Paris were originally dug, there was nothing under the streets like we find today. Sewers and water pipes, yes, but half of the gas mains in NY were installed after 1940, to say nothing of electricity and telecomms. Moving this, particularly while in use, is expensive. When building a skyscraper, you can assume that the cost of the electrical and mechanical subcontracts will each be larger that the costs of the entire structure (concrete and steel combined). That is for new construction, without the additional cost of keeping existing service running.

    There are many other conditions that can make TBMs cheaper. In hard rock, the cost of blasting from the surface could be more expensive than a main beam TBM (because you have to blast the over burden you are going to fill back in). On wet sandy soils, the cost of shoring and dewatering could be more expensive than a EPB TBM. Many cities have long straight boulevards, but many others have narrow chaotic streets like London. A city with a lot of hills cannot do cut and cover. Any water crossing requires a TBM or immersed tube (cannot cut and cover under a river!), and the cost of TBM launch boxes in the middle of the city for a few hundred meter crossing (or the effects on navigation/the environment of excavating the riverbed to set an immersed tube) may outweigh excavation savings on the rest of the line (TBMs have high overhead cost, so once you start digging it is best to dig as far as possible).

    If cut and cover isn’t being widely used anywhere there might be a reason. Around the world there are examples of transit agencies that get all kinds of things right (cost in Scandinavia and Spain, takt/pulses/POP in German speaking countries, subway/mainline integration in Japan, etc.); if not a single agency anywhere consistently builds cheap using cut and cover we should ask why. TBMs are not technological marvels for fun, they have advantages that should lower absolute cost (high automation, 24/7 operation, no need for separate shoring, dewatering, etc.). In most industries automated continuous production is always cheaper than batch production.

    There may be a good argument that developing countries with wide straight streets starting a system from scratch should consider cut and cover, as then can plan their lines shallow without existing lines, and because labor is be cheaper compared to the capital cost of TBMs. But this is a very specific case, far from a blanket “cut and cover is underrated.”

    • Herbert

      There is a downside to cut and cover that is sometimes brought up by transit geeks in the Nuremberg context:

      Many Nuremberg U-Bahn stretches between stations are built in such a way that the lowest point is between stations – thus the train gets a boost when starting from the station rolling downhill and braking is made easier by going uphill towards the station. If you’re running a TBM, all you gotta do is point it slightly downwards or upwards to build that. If you’re doing cut and cover this whole thing becomes, well, not impossible, but a lot more difficult…

    • Mikel

      It seems that the relative cost of different tunneling methods varies by city and the specific urban conditions. In the L11 extension proposal in Madrid (dossier, schematics, cost estimate), these are the quoted market costs for each method (p. 39):
      -Bored with 8.43-meter diameter TBM: €10,650/m.
      -Mined with Belgian method: ~8m width x 6.5m height: €8,850/m.
      -Cover-and-cut, 7.8m width: €15,000/m plus utility relocation.
      Bear in mind that those are the “presupuesto de ejecución material” i.e. bare costs without VAT, overhead nor profit margin (you’d have to multiply the project total by ~1.44 to get the total cost estimate) but the relative difference is interesting — the alignment is rather deep (going under the river, the M-30 and several older lines) which I suppose skews the comparison towards TBMs.

      The chosen method is TBM for all but the shallowest parts, but the stations are cover-and-cut, oscillating between €30 million for Conde de Casal (interchange with L6) and €55 million for Atocha (interchange with L1 and Cercanías). For utility relocation (p.76, 154, 219), the cost oscillates between €0.58 million at Madrid Río version 2 and €4.21 million for Palos de la Frontera version 1. Same caveat applies.

      • michaelrjames

        @Mikel:
        “-Cover-and-cut, 7.8m width: €15,000/m plus utility relocation.”

        Too ambiguous. You should use “incl.” or “excl.” for inclusive/exclusive of the utility relocation cost. You cite this as the highest cost so I assume you meant inclusive?

        • Mikel

          Oh, apologies, I meant exclusive — the tunneling cost is just for digging the tunnel, nothing else. The main utility relocation items are (p.48):
          -Sewers: visitable €1,500/m; large-diameter visitable €6,000/m; non-visitable €120/m.
          -Electricity: high voltage lines €600/m; low and medium voltage lines €400/m; public lighting €100/m.
          -Telecom: €500/m.
          -Gas lines: €280/m.

          • michaelrjames

            @Mikel

            Tx for clarification.
            So cut-and-cover is the most expensive–just for the trench itself without any utilities relocation etc! Is it because it is very deep, to match the TBM-bored bits? But then why not use the TBM for these bits too? Perhaps the trenched bits are just short sections compared to the bored bits?

          • Mikel

            @michaelrjames yep, the current L11 terminus at Plaza Elíptica is very shallow because it crosses above the older L6 tunnel, so the section between that and the TBM pit will be done cover-and-cut (schematics p.24). However the cul-de-sac after the future terminus at Conde de Casal will be mostly mined despite going under a wide avenue, presumably because ripping up the M-30/A-3 interchange would be too disruptive.

            The dossier also discusses the pros and cons of using TBMs or the traditional Madrid method (p. 41-42). The TL;DR is that using a TBM has a high overhead in time and money but also a much higher yield, which for a long tunnel results in a lower overall cost and a shorter timeframe.

    • Matthew Hutton

      Land was also much cheaper than it is today when the district line was built.

  10. Martin

    Alon, I’m curious about your thoughts about cut & cover for LA’s purple line. Despite the station being built as cut and cover, you wouldn’t know it when driving as seen at 6:30 in this video: https://youtu.be/CA6dUkxR9Ho

    This seems about as unintrusive as using a TBM.

  11. Frederick

    Two reasons why Bangkok chose to bore a tunnel instead of cut-and-cover:

    1. Bangkok has a notorious traffic problem. It is very undesirable to block major boulevards like the Rama IV Road for any amount of time.

    2. Bangkok is a rainy city but notoriously bad in water drainage infrastructure. This is why you don’t see many road underpasses in Bangkok (meanwhile, overpasses are plenty in Bangkok). Tunnel boring machines can work in a rainy day, can drill in waterproof hard rock, can deploy watertight tunnel shields while drilling, and the tunnel drilled will have a fixed geometry so engineers can easily predict where the water will go.

  12. wiesmann

    One datapoint that could be relevant is the building of the Löwenstrasse part of the Zürich main-station. My understanding is that it was kind of cut and cover, concrete structures (wall and ceiling) were built and then the space hollowed out.

  13. borners

    Umm sorry to sound look like the no-nothing here, but I’ve struggled to find materials detailing where and when you can use cut-and-cover. E.g. basic stuff on how wide a street needs to be for cut-and-cover to work? Or how cut-and-cover strategy deals with obstacles (e.g. rivers) where tunnel boring is necessary. How much does that screw up the cost advantage?

    • df1982

      I’m no expert, but my impression is that once you go to the trouble of setting up a TBM (purchasing the equipment, acquiring land for and then setting up a tunnel launch site, etc.), then it is relatively cheap to just let the machine keep on digging. This makes the economics of doing a line partly as cut-and-cover and partly as deep bore not stack up as well: so you either do the whole thing as cut-and-cover if you can or if not (which is more likely) then you end up having to do it all as a deep bore (this is excluding the stations, which are treated differently for construction purposes). Combined with the political issues of property resumption and traffic disruption from cut-and-cover, this probably explains the comparative rarity of cut-and-cover in modern subways. Governments will always be willing to throw more money at a project if it means copping less political heat.

      • Henry Miller

        It is more complex than that.

        A TBM has a maximum rate of progress. If you decide you want to go faster than that you either need to buy a second TBM, or you TBM where you must (plus a little more?), and cut-and cover the rest. For cut-and-cover any city has a large set of excavation companies who are experienced and have the equipment, you can hire as many of them as you want to work on different sections at the same time.

        TBMs only work in limited soil conditions. If you are going under a river (common in most cities and must be a TBM) that is one type of TBM. As you go farther in the city you may discover that your TBM won’t work well in that soil. There are combination TBMs, but you need to get the right one at order time, and they are more expensive. IIRC London cross-rail used 6 TBMs, two only for short under river segments. There is potential to TBM only in soils where it is needed and cut and cover the rest.

        Cut-and-cover is much shallower than TBM. Users will be much happier with a short walk down to the cut and cover station vs the much longer ride to the deeper TBM station, because off the overall time saved getting to the train. For long distance trips this doesn’t matter, but a cut and cover station is better for shorter trips (if you very short want hop-on-hop off you still need to be at grade).

        As an aside to that, while elevators are needed for the disabled you can save a lot of money making normal stations access a cheap stairway, while the deeper TBM station demands a more expensive escalator. This is additional station savings and worth exploring. (I believe in busy stations escalators handle more people and so are wroth it anyway, but for less busy stations)

        A used TBM may have value when you are done if you don’t use it too much. Thus if you only need to TBM a short stretch it is worth looking at what the used market is like. This limits you to what is on the used market (your tunnels might be a different size than ideal). Or more likely you can hire a TBM and crew from a company that works world wide. Yes you will pay top dollar per meter of tunnel with this plan (setup/teardown of a TBM isn’t cheap), but since you are only paying for a short section it can still be cheaper – have your accountant run numbers.

        I don’t know a TBM that is already digger compared to cut and cover in costs though. That is a number well worth considering with the above.

        • borners

          Thanks for these replies.

          So sticking with London as an example it wouldn’t be feasible to do say the Bakerloo line extension with cut and cover down the Old Kent road at least to New Cross station then TBMing it past Deptford? Or extend the Northern from Battersea to Kingston along the A3?

          • Henry Miller

            Looking at it from my home in the middle of the US it seems possible. However this is a case where location details matter a lot. I’ve never been to London, so I really don’t know what the reality is.

            Many streets are too narrow. If you can’t fit the trains on the streets (on some mythical day where no cars or people used it), you can’t fit them underneath it. TBM can go under buildings (though you need to take care) so there is more space to work with. Corners are a big issue here, even if the trains fit width wise, sometimes the corners are too sharp for the long train cars.

            What else is underground? London has rivers that no longer see daylight: they were paved over years ago. Utilities have similar issues, though they are easier to relocate in general. The above may not appear on modern maps, and even when they do in many cases the route they run is only an approximate guess as nobody has verified how they go between known points.

            The above is what I personally know about the situation in London. If the people doing the planning don’t know a lot more than me about the local situation something is seriously wrong. (I might disagree with their decisions if I knew as much as them, but right now I don’t know enough to make a valid comment)

          • fjod

            Yes I think this is probably totally feasible to engineer. No idea what’s under the Old Kent Road but it isn’t rivers. You’d need a TBM at the northern end though, especially if you plan to realign Elephant & Castle station (which is in the current plans) – no idea if this would make the cost savings negligible. I also think it may be politically feasible, seen as many/most users of the Old Kent Road would benefit from the Bakerloo line extension.

        • Onux

          @Henry Miller

          TBMs have a fixed rate of progress, but this is almost never an issue. Hiring more companies to dig C&C would be faster, but it would also mean you have to pay the cost of excavation all at once. For many public works projects budgeted over several years, this isn’t an option (just because the legislature will give you $1B/yr for five years doesn’t mean they have $5B to give the first year, to oversimplify). What’s more, even if you have lots of excavation companies, you won’t have lots of companies that can install track, signals, etc. Digging 10km of tunnel in a year doesn’t help if you then install 2km of track per year; if track install begins at one end and runs to the other anyway better to just let it follow the TBM. Soil types across the whole route will always be known before beginning to dig anyway.

          Some TBMs are only useful in one type per soil (i.e. main beam TBMs can only work in hard non-porous rock) but others can handle a variety of soils (i.e. earth pressure or full shield TBMs).

          Cut and cover can be very shallow (a la the Paris metro) but for most purposes the minimum depth for a TBM is not an issue, if becomes the mezzanine. If you have to go under hills or existing lines you are going deep regardless of method, which is probably why most tunneling is by TBM these days.

          Stair vs escalator is a negligible cost in building a whole subway; as noted above your choice of depth across the whole line may be determined by other factors even if you want shallow stations.

          • Henry Miller

            @Onux

            For the most part I will agree with your considerations. There are a lot of trade offs.

            The only exception is cost: The legislator shouldn’t be giving you $1billion/year. They are going to borrow the money, so $5 billion in one year or $5 billion over 5 years isn’t much difference if you can build that fast. Note that I said shouldn’t, that is a very different from what they actually do in practice. Though as we all know elections happen and that often results in projects being canceled.

            Doing 5 separate back to back $1billion projects makes a lot more sense than a 5 year project though. (I suspect you can’t get the time down to less than 1.5 years, but a less than 1 year goal would be good). Year one gets you a maintenance shed and one stop that is in the cheap area where you do maintenance. Year two gets you an extension that starts to be useful. One thing I’ve learned in computers is you can waste a lot of time planning without getting anything done, even if the whole is a 5-10 year project you should have good milestones along the way, and regular service (even if useless service) is a great milestone that can’t be taken away easily in the next round of cuts.

  14. Pingback: News roundup: why transit is expensive – Seattle Transit Blog
  15. Walter

    Although Cut-and-Cover appears to have greater disruption on the surface – I am not convinced that is the case. Based on Toronto, Canada, the TBM comes through with the tunnel. Then at each station, which is 25 to 30m deep on average, they cut down and build the station. This station construction, because of depth, takes about 4 or 5 years. This long lasting and disruptive construction happens at the intersections, where the stations tend to be, so the disruptive effects are amplified.
    For cut-and-cover, the entire line would need to be dug from above. But, for most of the length, it’s a simple matter of digging down maybe 10-12m, suspending some utilities, placing precast tunnel segments, and covering things up. This construction is measured in months (2 or 3) and not years. Then when you get to the stations, which we described as the most locations for disruption, the construction can take 1/3 less time since the stations are so shallow.
    Let’s assume station platforms are 150m long and spaced at 750m. What is worse;
    1) 4.5 year construction for 20% of the length (at critical intersections), or
    2) 3 year construction for 20% of the length (at critical intersections), plus 3 month disruption for the remaining 80% of the length?

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.