Despite TBMs being used all the time it is genererally much cheaper to build a tunnel by cut and cover - that is just call in excovators and dig a hole then cover it. when an excavator breaks you push it aside and get another until it is fixed. Want to go faster - just hire more excavators the overall cost is the same since you are done faster.
tbms are cheaper if you must go deep, but deep is a negative for subways since people must get down which takes time (faster is always better).
It is claimed that cut and cover is toa much surface disruption - but cut and cover is much faster and so the disruption to any one place can be quick enough. Better to manage disruption than eliminate it.
when a tbm breaks your project is delayed while you wait for parts and then fix it in place.
Tbms do go under everything else, but experience proves you can find things and go around them quick enough (sometimes there is a delay as somethingeof archeological interest is found - but evperience shows this is still cheaper than a TBM.
Yes TBMs are cool and useful but the simple shoud not be over looked
You're often not just building under existing things, but through them. We now have a lot under the roads. The Victorians built the subsurface lines of the London Underground with cut and cover, but Oxford is currently suffering overrunning works to lower a road under the station, because of unknown brick arches and utilities. Even the TBMs are building beside existing tunnels and basements.
What projects in developed countries have used cut and cover recently? In trying to find out, I see that HS2 under west London and the Canada line under Vancouver chose tunnels over cut and cover because it was cheaper.
> I see that [...] Canada line under Vancouver chose tunnels over cut and cover because it was cheaper.
Canada Line was mostly Cut-and-Cover - only the bits below downtown and crossing below the water were bored, the bulk of the underground was done cut and cover for cost and speed to make sure it opened for the 2010 olympics.
It was not a popular choice - not really announced before the project was approved, and local businesses along the route took a big hit.
Vancouver's current Broadway Line Extension is being done with TBMs to avoid the impact that the cut and cover canada line segment construction had.
> What projects in developed countries have used cut and cover recently?
Not really a new project, but parts of the subway in Stockholm are cut and cover. One of those tunnels (from the 1930s) has been leaking in water for some years and is up for a total overhaul, so basically digging up everyting and doing a new cover.
The section is 8 m wide and 925 m long, projected timeline is 6-7 years starting this fall. It will be a massive project, as one of the busiest streets in Stockholm is directly on top of it.
The article said out of 89 current projects 80 were TBM so it isn't surprising you don't know of the exceptions. I don't off the top of my head either.
I imagine it's also a bit difficult to separate it this cleanly, as most bigger projects will probably use a mix of technologies: cut and cover where possible (if it leads to savings), TBMs or other technologies like NATM (https://en.wikipedia.org/wiki/New_Austrian_tunneling_method) for the rest. Even if TBMs are used for the tunnels, cut and cover will probably be used for things like stations, emergency access points and intermediate TBM starting points (of course, the TBM starting points might be future stations).
Combining is an option. However a large part of the cost of TBM in the initial get it into place and then when you are done taking it out (sometimes you just leave that expensive machine down there). Thus if you must use a TBM the farther you can go in that one dig the overall cheaper the tunnel.
My neighborhood in Seoul got a new subway line while I grew up, I believe they used cut-and-cover (it was almost at the outskirt, so they didn't have anything else underground) and it took several years.
So, while it's doable, it's not exactly "Just dig a hole, build the tunnel and cover, done." It's a major disruption to shops along the street.
It's not a choice of TBM vs cut-and-cover. You also have drill-and-blast. Which method is better depends on a number of factors, including depth and the state of the soil/rock you're going through.
A problem with TBMs is it can be difficult/costly to secure the tunnel sufficiently against water ingress. If you mess up that one the cost of your tunnel can easily double or more.
Here in Norway we recently built a railway tunnel with TBMs (Blixtunnelen) and it's having problems with water ingress. A fairly mild problem relative to how bad these can get, but it's enough that the tunnel constantly has to be closed for repairs to the railway infrastructure due to water drips.
There are many options with different compromises. Cut and Cover comes in many different versions (some of them you build the cover first and then dig under that). However I'm going to double down and suggest that despite the disadvantages everyone else has raised the overall much lower cost of cut and cover, combined with the advantages of a shallower tunnel, makes it the right answer for a lot more situations. Most people claiming to need something else for some local feature are wrong - they can work around that other objection and be better off. Note that I'm not saying the objection is invalid, just that it should be worked around.
They are the political option. They disrupt the surface the least and so there are less political objections even though they cost more around the world.
Tunneling engineers do what the customer says. They will tell you what the options are and what they cost then let the customer choose the evpensive option if they want. Their job isn't to choose between options it is to eliminate the impossible (unsafe) ones and let the customer choose the pros andecons of the rest.
Cut and cover only works if your subway follows streets. Yeah, I think we should do that where it makes sense but quite often you need to go under buildings or hills.
Most of the time your subway will follow streets though. Your city has built around the existing streets and people have set their lives up around the places they can get to by the existing streets. If there are two points that are near each other as a crow flies but difficult to reach (shore to an island where there is no nearby bridge) people have made sure they don't need to go there and so you rarely gain much by giving them some new ability. While if you build on the streets they can now go places they were already going via your new subway.
Note that I said street and not highway. Some highways (limit access) are good for longer trips, but your subway would be better off using a side street parallel to the highway since while most people use the highway your subway doesn't need to get to any point on the highway while points on the side street - while of limited interest have at least some interest. City transit design is a complex subject that whole books are written about (sadly the people who are in charge don't read them)
That seems a very car-centric or American view, with cities built on a grid.
Older cities without existing grids can have older metro lines following streets (a limitation of the time they were built, by cut and cover) but newer ones generally don't follow streets. They provide more direct routes, and new crossings over rivers or to islands.
Look at Prague or especially Copenhagen for example.
The HS2 cut and cover tunnels were all in greenfield: https://www.hs2.org.uk/building-hs2/tunnels/green-tunnels/ ; much of the cost there goes on planning and documentation, an under-appreciated cost. It's also questionable as to whether they were needed at all; a plain cutting with embankment open to the air would have been fine from a civil engineering point of view, or even in many places just flat track, but the tunnels were planned because people objected to a railway running through fields.
Thing is, you’re often building these under already-existing things. The Dublin metro project is facing significant planning challenges on the basis that thousands of homes and businesses passed can complain about speculative vibrations from the TBM; with cut and cover it would just be a complete nonstarter (I think one small section is planned to be cut and cover, but most of it will be either bored or surface rail).
This is not cut and cover. Cut and cover is when you "cut" an excavation from the ground surface down and then "cover" it back up with dirt after constructing the structure you want below ground. You have linked to an article on NATM, which is a conventional (read: non-TBM) tunneling method.
Cut and cover is what I described: excavate down, build the structure, cover it up. I am aware that there are lots of options, you can browse my comment history and see that I probably have more experience designing and constructing underground infrastructure (and tunnels in particular) than anyone else that posts on HN.
Cut and Cover is often used in transit conversations to mean any variation where you dig from the surface and not the construction method that is cut and cover. These are two different domains talking about different things that are related. Something we both need to be more aware of in conversations.
Sometimes a temporary cover is built first to minimise disruption on the surface.
Here's an old example of an umbrella bridge over Oxford Circus during the construction of the Victoria Line. There's a longer video out there of the construction of the Victoria Line that covers this in more detail.
Some of the stations on the Broadway Subway Skytrain extension in Vancouver use a similar approach, where half the road is closed and a road deck is built, then traffic is shifted over to the new road deck while excavation takes place from the side. There's some great views of this while riding the bus.
This is not what the cover in cut and cover refers to. This is a temporary bridge erected over a cut and cover operation. When the excavation (the "cut") and structure are completed underneath and the backfill/concrete is placed up to ground surface (the "cover") the temporary bridge will be pulled out. The "cover" part is about covering up the buried structure after it's built to put the ground surface back to where it was.
Edit: If the road deck is left in place permanently then it is a permanent elevated road deck built over a cut and cover tunnel. I can see how some people might consider this the "cover," but that is atypical in the industry and not what people are usually talking about when they say cut and cover. I'll concede that this approach sometimes happens, but I wouldn't call it "often" like GP does and I'd also note that even under this scheme the final surface/cover (e.g., the roadway) is completed after the underground excavation and structure are finished, meaning that the cut still precedes the cover.
The most common terms for this in the industry are "lid" or "cap." As in, you put the lid on, or cap, the excavation or cut and cover tunnel.
Some time ago I saw a list of all (i'm not in construcion so I only assume it was all) the different options. many people have innovated many different ways to dig while minizing surface impact. For my purposes if you eventually dig down I consider them all cut and cover - but of course if you are in construction the differences matter. For that matter even to me chose matter - but only as details that we need to argee on before starting work.
TBMs are for going around problems. Compared to cut and cover you can ignore entire classes of people who would be, expensive, delaying obnoxious stakeholders.
<gestures at the contents of an average HN comment section whenever the subject is public infrastructure>
Do you have any interesting examples of modern day cut and cover that are not part of a TBM run?
> TBMs are cool and useful
They are what is needed to move humanity forward.
You have trillion $ centers that using 3D you can add massive amounts of access to. This is not like reclaiming land from water, which is ~stealing, this is value adding through topography, it's creation.
The tunnel will be around for 100 years. Not much ophortunity cost is lost. And you lose the ophortunity cost of what ever you would spend the saved money on for ever. A longer line for example.
I want to mention a recent extension to Berlin‘s subway system right in the middle of the city, in swampy ground and below a canal. They injected „freeze lances“ into the ground, turned 43,000 m^3 of mud into an ice block and then carved their way through it. Absolutely amazing engineering, I think planned and executed by a Swiss company.
So it doesn’t seem to be mentioned here, but has there been some major recent change in the economics of the _scale_ of the tunnels? Dublin’s beleaguered metro project recently put out its new design; there are a bunch of changes from the previous early-noughties design, but one of the more eyebrow-raising ones is that the bored sections will now be a single _9.5 meter diameter_ tunnel; it is claimed that this is less disruptive and cheaper than the two-narrow-tunnel approach used in conventional bored metros. Recent metro lines is Barcelona and Madrid have also used the single-giant-tunnel approach.
Anyone know what’s going on here? Intuitively you’d expect two small tunnels to be cheaper; you’re moving a lot less material.
I suspect the cost is not in volume of material removed, but in operation and maintenance of the machine and it's crew. One tunnel instead of two is half as much machine and crew time.
Please explain how you reasoned this out. If one 9.5m tunnel requires 70 m³ of removal for every meter of forward progress, and two 5m tunnels requires 40 m³ of removal for every meter of forward progress, won't it take nearly twice as long to dig the single larger tunnel? Won't that require twice the machine hours and twice the crew hours?
How does the volume of material removed not directly impact machine and crew costs?
Two dump trucks can't pass eachother in a 5m tunnel. That'll put a restrictive upper bound on material removal rate which isn't present in the larger tunnel.
Let's say the larger TBM can remove 7/4 the debris in the same amount of time. That means it uses 7/4 more carts or has 7/4 bigger carts and uses 7/4 more energy to power those carts. It places 7/4 more wear on the rails and has a higher probability of increased maintenance costs on those rails, let alone the TBM itself. All of this to make the same forward progress in the same amount of time.
Call me stubborn, but I need a better theory to explain why the bigger TBM would be cheaper. You don't get to pretend that larger material removal rates are free. You don't get to pretend that the 7/4 more cutting wheels don't cost more money to replace. The 7/4 more energy to run the TBM isn't free.
Yes, the larger machine would have to have larger material removal rates just to keep up with the smaller machine. But that makes it more expensive, not less.
> Yes, the larger machine would have to have larger material removal rates just to keep up with the smaller machine. But that makes it more expensive, not less.
More expensive than one, potentially less expensive than two. Happens in manufacturing and industry all the time. Transport cost alone could be enough to account for it.
> Call me stubborn, but I need a better theory to explain why the bigger TBM would be cheaper.
Sounds like you should have a chat with the project lead.
One of the references in the article mentions using better alloys ("modified tool steel") and 20" discs produces better results and lower overall costs due to time between disc replacement. This could be A8 "chipper knife" steel.
Makes sense. I was thinking more about the face of the machine itself being larger, likely the size of the tunnel, and would contain many such 20" discs. In other words, given that the material removal rate of each disc is likely fairly similar from machine to machine, the machine with more of them removes more material at once.
It seems self-evident that all additional material removed has 0 additional value. In fact it has negative value because it increases the requirements for any support or finish materials. So energizing and consuming 7/4 more discs is just pure waste.
It also adds disposal costs. The most obvious argument I can think of to support this approach is if the excess material could be sold at a profit. But that seems dubious.
> So energizing and consuming 7/4 more discs is just pure waste.
That is and assumption and not self evident. There are lots of reasons one might want to use many cutters. It is a common practice in wood and metal machining, for instance[1]. Having additional cutting surfaces reduces the rate of wear, provides redundancy in case of chipping, increases the mean time between required services, allows cutters to run cooler, reduces chatter, improves surface finish, and a host of other reasons.
You seem to have focused your 'tunnel vision' on the costs of boring that thing alone.
Sort of CAPEX, so to speak. But what about less OPEX later, because of more flexibility, less maintenance burden, when other demands arise or conditions change?
If the tunnel is wide enough, two tracks can be built side by side, and another level of two tracks on top of it. That makes room for a total of four (potential) tracks, with the possibility to switch between them not only horizontally, but also vertically via ramps (inside that tunnel profile). It also enables single platform stations for one track on one or two of the two possible levels in such a tunnel. Without any further 'big digs', or blocking the line altogether.
The larger TBM would require ~7/4 more discs. So you might go a longer time between disc changes, but each disc change requires 7/4 more effort and material. Any hypothetical savings in spin-up and spin-down would be offset. If the balance came out to some advantage, it would not be due to a larger diameter of the TBM. It would just indicate that you should consider retrofitting the smaller diameter TBM with thicker discs.
Two of everything, ventilation, electrical, water,
telecoms,you name a system, and then name it again.
Chances are that set backs required for engineering and existing infrastructure will be
challenging, with two parallel sytems.
edit: there is an assumption that the space required undergroud is free and infinite, when in fact two tunnels with set backs and spacing will require more total space than one larger tunnel
and will then limit the availible area for deep building foundations and anything else.
Then we get to interchandes, and emergency access.
And there is a ready market for any material removed from the dig, which will be a very routine operation.
And while the volume of fill removed will be greater, the internal volume per cubic foot of tunnel wall will also be greater, allowing for more generous alotment of space for future systems and untilities, bigger wider lanes, etc.
Somewhere I came across a journal, just for tunnel makers, and there equipment.
And recent reports have tunnels bieng completed
ahead of schedule and on budget, so they are infact, taking care of business.
Oh, yep, I’ve no doubt that if they can ever actually get planning through for the damn thing, they’ll be able to build it. In Ireland it’s generally the planning that’s the problem.
What I’m curious about, though, is what, if anything, changed to make this single-bore approach suddenly popular. A lot of new metro lines seem to be built that way, but old (and not, like, 19th century, even most 20th century stuff) always seem to be double-bore.
Has anyone attempted to use water jets with aggregate, like you would on a CNC waterjet cutter, for boring through rock?
Had this thought the other day and wondered if there would be any advantages over metal cutting teeth. From the article, I gather that replacing cutting teeth isn't that much of a problem nowadays so that mitigates one advantage it might have.
Interesting article. This bit was not what I was expecting:
Today, rock tunneling machines have achieved tunneling rates of over 700 meters per week, and soil tunneling machines have achieved rates of over 200 meters per week, though this is dependent on the size of the TBM
Ie, that rock tunneling can be > 3x quicker than soil tunneling.
It never seems to get stated directly by tunneling through stable rock has to be really easy. Like if you can drill and blast your way through you just need a small crew of workers and a dump truck.
There is a half mile long tunnel in the California desert that an addled miner made with hand tools and dynamite over 38 years. Self funded and part time.
While we're talking about cool boring technologies... I once heard that Seoul's Bus Terminal Station, line 9, was a particularly difficult case. Line 9 had to go beneath existing line 3, but it couldn't go too deep because it (obviously) needed to connect to the rest of the line.
In the end they thrust a group of horizontal steel pipes to stabilize the soil and dug beneath it - and those pipes were 15 cm(!) below existing line 3.
I wonder how far off we are from vaporizing the material and venting it out of the tunnel.
Probably way more expensive but you could go faster and don’t need to deal with hauling the material out. It might glassify the surrounding soil and make its own support structure too.
This topic has been mentioned in regards to digging super deep bore holes for geothermal energy. The high temps a couple miles down started to soften metal drill bits but some companies are working on lasers that are capable of melting rock to allow further digging.
An 80W CO2 LASER is plenty of power. I regularly engrave the living shit out of stone with mine, and that goes a few millimeters deep.
Get a couple kW in the same wavelength (10640nm) with a proper adjustable focus (movable diopter) and a galvo and a vacuum assist and you can just start raster-scanning your way through the rock.
Yes, they have been. The growing tip of a root is called the root apical meristem[1] but the short story is that they work effectively by being tiny and taking years or decades to develop, so not all that helpful in terms of making things easier or cheaper. They also can't really go through rock, though they can often split them, particularly if they have fractures already.
I don’t believe they have advanced the field in any way. They didn’t really have any solid ideas, just the idea that it would be really cool if tunnelling could be made cheaper. Surprise surprise, it’s actually a highly advanced, mature field and there wasn’t actually any low-hanging fruit for inexperienced startup guys to disrupt…
The only thing they’ve done with their Vegas tunnel is to make it cheaper by making it too small to have proper emergency egress, but they can get away with it since with just their chauffeured Teslas going through it only fitting a few people each (making the system extremely low-capacity), they can’t really fit many people in at once, and it’s short enough so combined that makes it not that much of an issue.
In the newest section of the Vegas tunnel they have now only built a one-way tunnel instead of a tunnel for each way, making it cheaper again but low-capacity to a basically comedic level.
Is there any reason to think they will be able to? They are just going the same thing as everyone else but started later. Another player might lower costs slightly but not much.
The strip is 4 miles long… when Musk company can bore longer, wider tunnels avoiding other infrastructure e.g. London tube lines, sewers etc then they might have something to shout about
I believe he is deliberately targeting low process burden, low soil resistance targets. He isn't aggressively chasing contracts for complex, deep, large tunnels working around existing infra, or rail, he's chasing places which want disruptive change in transport.
Nobody has explained to me what the boring company has actually done innovating TBM, the machines themselveds distinct from a cash injection and some energy from Musk as lead investor. TBM are flow process. It's a pipeline of inputs and outputs. It's scheduling.
The LV tunnel isn't a complete success either. It's not operating completely as promised yet is it?
Despite TBMs being used all the time it is genererally much cheaper to build a tunnel by cut and cover - that is just call in excovators and dig a hole then cover it. when an excavator breaks you push it aside and get another until it is fixed. Want to go faster - just hire more excavators the overall cost is the same since you are done faster.
tbms are cheaper if you must go deep, but deep is a negative for subways since people must get down which takes time (faster is always better).
It is claimed that cut and cover is toa much surface disruption - but cut and cover is much faster and so the disruption to any one place can be quick enough. Better to manage disruption than eliminate it.
when a tbm breaks your project is delayed while you wait for parts and then fix it in place.
Tbms do go under everything else, but experience proves you can find things and go around them quick enough (sometimes there is a delay as somethingeof archeological interest is found - but evperience shows this is still cheaper than a TBM.
Yes TBMs are cool and useful but the simple shoud not be over looked
You're often not just building under existing things, but through them. We now have a lot under the roads. The Victorians built the subsurface lines of the London Underground with cut and cover, but Oxford is currently suffering overrunning works to lower a road under the station, because of unknown brick arches and utilities. Even the TBMs are building beside existing tunnels and basements.
What projects in developed countries have used cut and cover recently? In trying to find out, I see that HS2 under west London and the Canada line under Vancouver chose tunnels over cut and cover because it was cheaper.
> I see that [...] Canada line under Vancouver chose tunnels over cut and cover because it was cheaper.
Canada Line was mostly Cut-and-Cover - only the bits below downtown and crossing below the water were bored, the bulk of the underground was done cut and cover for cost and speed to make sure it opened for the 2010 olympics.
It was not a popular choice - not really announced before the project was approved, and local businesses along the route took a big hit.
Vancouver's current Broadway Line Extension is being done with TBMs to avoid the impact that the cut and cover canada line segment construction had.
It also resulted in a rail line that has to slow down significantly to round screeching curves along the path of the road above.
> What projects in developed countries have used cut and cover recently?
Not really a new project, but parts of the subway in Stockholm are cut and cover. One of those tunnels (from the 1930s) has been leaking in water for some years and is up for a total overhaul, so basically digging up everyting and doing a new cover.
The section is 8 m wide and 925 m long, projected timeline is 6-7 years starting this fall. It will be a massive project, as one of the busiest streets in Stockholm is directly on top of it.
> What projects in developed countries have used cut and cover recently?
I know in Sweden the Västlänken project partially used cut-and-cover at least for the part going southwards..
https://en.wikipedia.org/wiki/West_Link
The article said out of 89 current projects 80 were TBM so it isn't surprising you don't know of the exceptions. I don't off the top of my head either.
I imagine it's also a bit difficult to separate it this cleanly, as most bigger projects will probably use a mix of technologies: cut and cover where possible (if it leads to savings), TBMs or other technologies like NATM (https://en.wikipedia.org/wiki/New_Austrian_tunneling_method) for the rest. Even if TBMs are used for the tunnels, cut and cover will probably be used for things like stations, emergency access points and intermediate TBM starting points (of course, the TBM starting points might be future stations).
Combining is an option. However a large part of the cost of TBM in the initial get it into place and then when you are done taking it out (sometimes you just leave that expensive machine down there). Thus if you must use a TBM the farther you can go in that one dig the overall cheaper the tunnel.
My neighborhood in Seoul got a new subway line while I grew up, I believe they used cut-and-cover (it was almost at the outskirt, so they didn't have anything else underground) and it took several years.
So, while it's doable, it's not exactly "Just dig a hole, build the tunnel and cover, done." It's a major disruption to shops along the street.
It is a major disruption for a couple years.
It's not a choice of TBM vs cut-and-cover. You also have drill-and-blast. Which method is better depends on a number of factors, including depth and the state of the soil/rock you're going through.
A problem with TBMs is it can be difficult/costly to secure the tunnel sufficiently against water ingress. If you mess up that one the cost of your tunnel can easily double or more.
Here in Norway we recently built a railway tunnel with TBMs (Blixtunnelen) and it's having problems with water ingress. A fairly mild problem relative to how bad these can get, but it's enough that the tunnel constantly has to be closed for repairs to the railway infrastructure due to water drips.
There are many options with different compromises. Cut and Cover comes in many different versions (some of them you build the cover first and then dig under that). However I'm going to double down and suggest that despite the disadvantages everyone else has raised the overall much lower cost of cut and cover, combined with the advantages of a shallower tunnel, makes it the right answer for a lot more situations. Most people claiming to need something else for some local feature are wrong - they can work around that other objection and be better off. Note that I'm not saying the objection is invalid, just that it should be worked around.
Are you a tunneling engineer?
That over 90% of projects are using TBMs strongly suggests they're usually the better option.
They are the political option. They disrupt the surface the least and so there are less political objections even though they cost more around the world.
Tunneling engineers do what the customer says. They will tell you what the options are and what they cost then let the customer choose the evpensive option if they want. Their job isn't to choose between options it is to eliminate the impossible (unsafe) ones and let the customer choose the pros andecons of the rest.
Cut and cover only works if your subway follows streets. Yeah, I think we should do that where it makes sense but quite often you need to go under buildings or hills.
Most of the time your subway will follow streets though. Your city has built around the existing streets and people have set their lives up around the places they can get to by the existing streets. If there are two points that are near each other as a crow flies but difficult to reach (shore to an island where there is no nearby bridge) people have made sure they don't need to go there and so you rarely gain much by giving them some new ability. While if you build on the streets they can now go places they were already going via your new subway.
Note that I said street and not highway. Some highways (limit access) are good for longer trips, but your subway would be better off using a side street parallel to the highway since while most people use the highway your subway doesn't need to get to any point on the highway while points on the side street - while of limited interest have at least some interest. City transit design is a complex subject that whole books are written about (sadly the people who are in charge don't read them)
That seems a very car-centric or American view, with cities built on a grid.
Older cities without existing grids can have older metro lines following streets (a limitation of the time they were built, by cut and cover) but newer ones generally don't follow streets. They provide more direct routes, and new crossings over rivers or to islands.
Look at Prague or especially Copenhagen for example.
https://openrailwaymap.org//mobile.php?style=standard&lat=55...
Streets have existed since cities began is prehistory times. even before cars people needed to get around.
Cut and cover is basically unusable in cities because you'd have to "cut" down all the buildings you'd otherwise tunnel under.
There isn't some industry saying "yes we could cut and cover here, but we prefer the slower more expensive option of a TBM"!
(I see NATM mentioned; there have been safety issues https://www.hse.gov.uk/pubns/natm.htm )
The HS2 cut and cover tunnels were all in greenfield: https://www.hs2.org.uk/building-hs2/tunnels/green-tunnels/ ; much of the cost there goes on planning and documentation, an under-appreciated cost. It's also questionable as to whether they were needed at all; a plain cutting with embankment open to the air would have been fine from a civil engineering point of view, or even in many places just flat track, but the tunnels were planned because people objected to a railway running through fields.
Thing is, you’re often building these under already-existing things. The Dublin metro project is facing significant planning challenges on the basis that thousands of homes and businesses passed can complain about speculative vibrations from the TBM; with cut and cover it would just be a complete nonstarter (I think one small section is planned to be cut and cover, but most of it will be either bored or surface rail).
That is a poor excuse - places with much worse problems have don cut and cover.
More information on cut and cover: https://de.m.wikipedia.org/wiki/Neue_%C3%96sterreichische_Tu...
(better translate from the german version, as it has more information and examples)
This is not cut and cover. Cut and cover is when you "cut" an excavation from the ground surface down and then "cover" it back up with dirt after constructing the structure you want below ground. You have linked to an article on NATM, which is a conventional (read: non-TBM) tunneling method.
I see; you are competely right, I am not really familiar with the the technical terms and messed up with just literally translating "cut and cover".
Cut and cover often is cover than dig - there are a lot of options
Cut and cover is what I described: excavate down, build the structure, cover it up. I am aware that there are lots of options, you can browse my comment history and see that I probably have more experience designing and constructing underground infrastructure (and tunnels in particular) than anyone else that posts on HN.
Cut and Cover is often used in transit conversations to mean any variation where you dig from the surface and not the construction method that is cut and cover. These are two different domains talking about different things that are related. Something we both need to be more aware of in conversations.
Cut and cover is what I said. Your definition (cover then dig) makes no sense. Cover what? If you haven't dug first then there's nothing to cover.
Sometimes a temporary cover is built first to minimise disruption on the surface.
Here's an old example of an umbrella bridge over Oxford Circus during the construction of the Victoria Line. There's a longer video out there of the construction of the Victoria Line that covers this in more detail.
https://youtu.be/9ghWdnYtp2I
Some of the stations on the Broadway Subway Skytrain extension in Vancouver use a similar approach, where half the road is closed and a road deck is built, then traffic is shifted over to the new road deck while excavation takes place from the side. There's some great views of this while riding the bus.
This is not what the cover in cut and cover refers to. This is a temporary bridge erected over a cut and cover operation. When the excavation (the "cut") and structure are completed underneath and the backfill/concrete is placed up to ground surface (the "cover") the temporary bridge will be pulled out. The "cover" part is about covering up the buried structure after it's built to put the ground surface back to where it was.
Edit: If the road deck is left in place permanently then it is a permanent elevated road deck built over a cut and cover tunnel. I can see how some people might consider this the "cover," but that is atypical in the industry and not what people are usually talking about when they say cut and cover. I'll concede that this approach sometimes happens, but I wouldn't call it "often" like GP does and I'd also note that even under this scheme the final surface/cover (e.g., the roadway) is completed after the underground excavation and structure are finished, meaning that the cut still precedes the cover.
The most common terms for this in the industry are "lid" or "cap." As in, you put the lid on, or cap, the excavation or cut and cover tunnel.
Some time ago I saw a list of all (i'm not in construcion so I only assume it was all) the different options. many people have innovated many different ways to dig while minizing surface impact. For my purposes if you eventually dig down I consider them all cut and cover - but of course if you are in construction the differences matter. For that matter even to me chose matter - but only as details that we need to argee on before starting work.
I believe the road decks on the Broadway Subway are permanent.
TBMs are for going around problems. Compared to cut and cover you can ignore entire classes of people who would be, expensive, delaying obnoxious stakeholders.
<gestures at the contents of an average HN comment section whenever the subject is public infrastructure>
> it is genererally much cheaper to build a tunnel by cut and cover
Coolo, you are building tunnels in a city that doesn't exist yet.
Start here - https://en.wikipedia.org/wiki/Opportunity_cost
Do you have any interesting examples of modern day cut and cover that are not part of a TBM run?
> TBMs are cool and useful
They are what is needed to move humanity forward.
You have trillion $ centers that using 3D you can add massive amounts of access to. This is not like reclaiming land from water, which is ~stealing, this is value adding through topography, it's creation.
I vouched for this comment, but you might want to check your general style of communication and then email dang about your apparent shadow ban.
The tunnel will be around for 100 years. Not much ophortunity cost is lost. And you lose the ophortunity cost of what ever you would spend the saved money on for ever. A longer line for example.
I want to mention a recent extension to Berlin‘s subway system right in the middle of the city, in swampy ground and below a canal. They injected „freeze lances“ into the ground, turned 43,000 m^3 of mud into an ice block and then carved their way through it. Absolutely amazing engineering, I think planned and executed by a Swiss company.
This is also a technique that was used more than 100 years ago for the line 4 of the Paris métro: https://www.ratp.fr/en/discover/out-and-about/culture/histor...
Funny that they had to freeze the ground to dig it, as it is one of the hottest line in Paris (at least 20 years ago).
My understanding is that the freezing is normally about stopping water intrusion long enough to drill and place concrete
Big Dig also used ground freezing. https://babelniche.com/2024/01/17/innovations-of-the-big-dig...
And the Erie Canal dig the sections through swampy ground in the winter. That was partially to avoid Malaria. Circa 1825
Got a link for this? I am interested :-)
I can find https://www.berlin.de/en/news/8097068-5559700-stabilization-...
Which states that due to sandy soil concrete was injected with lances. It was done by a French company.
Might be the same tunnel?
So it doesn’t seem to be mentioned here, but has there been some major recent change in the economics of the _scale_ of the tunnels? Dublin’s beleaguered metro project recently put out its new design; there are a bunch of changes from the previous early-noughties design, but one of the more eyebrow-raising ones is that the bored sections will now be a single _9.5 meter diameter_ tunnel; it is claimed that this is less disruptive and cheaper than the two-narrow-tunnel approach used in conventional bored metros. Recent metro lines is Barcelona and Madrid have also used the single-giant-tunnel approach.
Anyone know what’s going on here? Intuitively you’d expect two small tunnels to be cheaper; you’re moving a lot less material.
I suspect the cost is not in volume of material removed, but in operation and maintenance of the machine and it's crew. One tunnel instead of two is half as much machine and crew time.
Please explain how you reasoned this out. If one 9.5m tunnel requires 70 m³ of removal for every meter of forward progress, and two 5m tunnels requires 40 m³ of removal for every meter of forward progress, won't it take nearly twice as long to dig the single larger tunnel? Won't that require twice the machine hours and twice the crew hours?
How does the volume of material removed not directly impact machine and crew costs?
Two dump trucks can't pass eachother in a 5m tunnel. That'll put a restrictive upper bound on material removal rate which isn't present in the larger tunnel.
Note they use railway trucks to remove spoil (and transport everything else) from a TBM. A narrow gauge temporary track behind the TBM as it advances.
Seems like larger machines are likely to have larger material removal rates. Dependent on the surface of the digging tool and it's feed rate.
Let's say the larger TBM can remove 7/4 the debris in the same amount of time. That means it uses 7/4 more carts or has 7/4 bigger carts and uses 7/4 more energy to power those carts. It places 7/4 more wear on the rails and has a higher probability of increased maintenance costs on those rails, let alone the TBM itself. All of this to make the same forward progress in the same amount of time.
Call me stubborn, but I need a better theory to explain why the bigger TBM would be cheaper. You don't get to pretend that larger material removal rates are free. You don't get to pretend that the 7/4 more cutting wheels don't cost more money to replace. The 7/4 more energy to run the TBM isn't free.
Yes, the larger machine would have to have larger material removal rates just to keep up with the smaller machine. But that makes it more expensive, not less.
> Yes, the larger machine would have to have larger material removal rates just to keep up with the smaller machine. But that makes it more expensive, not less.
More expensive than one, potentially less expensive than two. Happens in manufacturing and industry all the time. Transport cost alone could be enough to account for it.
> Call me stubborn, but I need a better theory to explain why the bigger TBM would be cheaper.
Sounds like you should have a chat with the project lead.
One of the references in the article mentions using better alloys ("modified tool steel") and 20" discs produces better results and lower overall costs due to time between disc replacement. This could be A8 "chipper knife" steel.
https://www.robbinstbm.com/wp-content/uploads/2017/07/Smadin...
https://www.litechtools.com/products/tbm-disc-cutter/
Makes sense. I was thinking more about the face of the machine itself being larger, likely the size of the tunnel, and would contain many such 20" discs. In other words, given that the material removal rate of each disc is likely fairly similar from machine to machine, the machine with more of them removes more material at once.
It seems self-evident that all additional material removed has 0 additional value. In fact it has negative value because it increases the requirements for any support or finish materials. So energizing and consuming 7/4 more discs is just pure waste.
It also adds disposal costs. The most obvious argument I can think of to support this approach is if the excess material could be sold at a profit. But that seems dubious.
> So energizing and consuming 7/4 more discs is just pure waste.
That is and assumption and not self evident. There are lots of reasons one might want to use many cutters. It is a common practice in wood and metal machining, for instance[1]. Having additional cutting surfaces reduces the rate of wear, provides redundancy in case of chipping, increases the mean time between required services, allows cutters to run cooler, reduces chatter, improves surface finish, and a host of other reasons.
Seems like you could research it.
1: https://www.advancedmachinery.co.uk/content/product/8500_0_z...
You seem to have focused your 'tunnel vision' on the costs of boring that thing alone.
Sort of CAPEX, so to speak. But what about less OPEX later, because of more flexibility, less maintenance burden, when other demands arise or conditions change?
If the tunnel is wide enough, two tracks can be built side by side, and another level of two tracks on top of it. That makes room for a total of four (potential) tracks, with the possibility to switch between them not only horizontally, but also vertically via ramps (inside that tunnel profile). It also enables single platform stations for one track on one or two of the two possible levels in such a tunnel. Without any further 'big digs', or blocking the line altogether.
The larger TBM would require ~7/4 more discs. So you might go a longer time between disc changes, but each disc change requires 7/4 more effort and material. Any hypothetical savings in spin-up and spin-down would be offset. If the balance came out to some advantage, it would not be due to a larger diameter of the TBM. It would just indicate that you should consider retrofitting the smaller diameter TBM with thicker discs.
Two of everything, ventilation, electrical, water, telecoms,you name a system, and then name it again. Chances are that set backs required for engineering and existing infrastructure will be challenging, with two parallel sytems. edit: there is an assumption that the space required undergroud is free and infinite, when in fact two tunnels with set backs and spacing will require more total space than one larger tunnel and will then limit the availible area for deep building foundations and anything else. Then we get to interchandes, and emergency access. And there is a ready market for any material removed from the dig, which will be a very routine operation. And while the volume of fill removed will be greater, the internal volume per cubic foot of tunnel wall will also be greater, allowing for more generous alotment of space for future systems and untilities, bigger wider lanes, etc. Somewhere I came across a journal, just for tunnel makers, and there equipment. And recent reports have tunnels bieng completed ahead of schedule and on budget, so they are infact, taking care of business.
Oh, yep, I’ve no doubt that if they can ever actually get planning through for the damn thing, they’ll be able to build it. In Ireland it’s generally the planning that’s the problem.
What I’m curious about, though, is what, if anything, changed to make this single-bore approach suddenly popular. A lot of new metro lines seem to be built that way, but old (and not, like, 19th century, even most 20th century stuff) always seem to be double-bore.
One big factor for single-bore subway is that can fit the stations inside the tunnel. That saves a lot of money digging out the stations.
I think Barcelone Line 9 is the one that popularized it. I suspect the other factor is that they figured out how to make large TBM.
(2023)
Some more discussion then: https://news.ycombinator.com/item?id=37836889
Has anyone attempted to use water jets with aggregate, like you would on a CNC waterjet cutter, for boring through rock?
Had this thought the other day and wondered if there would be any advantages over metal cutting teeth. From the article, I gather that replacing cutting teeth isn't that much of a problem nowadays so that mitigates one advantage it might have.
Interesting article. This bit was not what I was expecting:
Today, rock tunneling machines have achieved tunneling rates of over 700 meters per week, and soil tunneling machines have achieved rates of over 200 meters per week, though this is dependent on the size of the TBM
Ie, that rock tunneling can be > 3x quicker than soil tunneling.
Rock often has sufficient structural integrity to get by with some bolting vs adding structure (steel, shotcrete, etc) to the soil tunnel as they go
It never seems to get stated directly by tunneling through stable rock has to be really easy. Like if you can drill and blast your way through you just need a small crew of workers and a dump truck.
There is a half mile long tunnel in the California desert that an addled miner made with hand tools and dynamite over 38 years. Self funded and part time.
https://en.wikipedia.org/wiki/Burro_Schmidt_Tunnel
Then digging through unstable rock with water intrusion and danger of flooding the costs have to go way way up.
While we're talking about cool boring technologies... I once heard that Seoul's Bus Terminal Station, line 9, was a particularly difficult case. Line 9 had to go beneath existing line 3, but it couldn't go too deep because it (obviously) needed to connect to the rest of the line.
In the end they thrust a group of horizontal steel pipes to stabilize the soil and dug beneath it - and those pipes were 15 cm(!) below existing line 3.
This ground support method is called spiling, for those interested.
I wonder how far off we are from vaporizing the material and venting it out of the tunnel.
Probably way more expensive but you could go faster and don’t need to deal with hauling the material out. It might glassify the surrounding soil and make its own support structure too.
This topic has been mentioned in regards to digging super deep bore holes for geothermal energy. The high temps a couple miles down started to soften metal drill bits but some companies are working on lasers that are capable of melting rock to allow further digging.
An 80W CO2 LASER is plenty of power. I regularly engrave the living shit out of stone with mine, and that goes a few millimeters deep.
Get a couple kW in the same wavelength (10640nm) with a proper adjustable focus (movable diopter) and a galvo and a vacuum assist and you can just start raster-scanning your way through the rock.
Has anyone studied how roots tunnel underground so effectively? Maybe we can copy aspects of that.
Yes, they have been. The growing tip of a root is called the root apical meristem[1] but the short story is that they work effectively by being tiny and taking years or decades to develop, so not all that helpful in terms of making things easier or cheaper. They also can't really go through rock, though they can often split them, particularly if they have fractures already.
[1]: https://en.wikipedia.org/wiki/Meristem#Root_apical_meristem
Nice text! I’m just surprised at the lack of mention of the techniques used in the Gotthard in the 19th century.
How is The Boring Company doing lately? There will be huge induced demand for tunnels if they manage to deliver even close to their goals.
I don’t believe they have advanced the field in any way. They didn’t really have any solid ideas, just the idea that it would be really cool if tunnelling could be made cheaper. Surprise surprise, it’s actually a highly advanced, mature field and there wasn’t actually any low-hanging fruit for inexperienced startup guys to disrupt…
The only thing they’ve done with their Vegas tunnel is to make it cheaper by making it too small to have proper emergency egress, but they can get away with it since with just their chauffeured Teslas going through it only fitting a few people each (making the system extremely low-capacity), they can’t really fit many people in at once, and it’s short enough so combined that makes it not that much of an issue.
In the newest section of the Vegas tunnel they have now only built a one-way tunnel instead of a tunnel for each way, making it cheaper again but low-capacity to a basically comedic level.
Is there any reason to think they will be able to? They are just going the same thing as everyone else but started later. Another player might lower costs slightly but not much.
They’re also boring much smaller tunnels which have much smaller costs
No reason, but Elon style is - if he delivers to deliver qualitive and not quantitative improvements to technology.
Slow and steady progress. They’re just about to go big in Vegas, with tunnels the length of the strip.
The strip is 4 miles long… when Musk company can bore longer, wider tunnels avoiding other infrastructure e.g. London tube lines, sewers etc then they might have something to shout about
I believe he is deliberately targeting low process burden, low soil resistance targets. He isn't aggressively chasing contracts for complex, deep, large tunnels working around existing infra, or rail, he's chasing places which want disruptive change in transport.
Nobody has explained to me what the boring company has actually done innovating TBM, the machines themselveds distinct from a cash injection and some energy from Musk as lead investor. TBM are flow process. It's a pipeline of inputs and outputs. It's scheduling.
The LV tunnel isn't a complete success either. It's not operating completely as promised yet is it?
[flagged]
Its a special "kind of fraud and failure" that launched 90% of mass to LEO in 2024.
"Success is the child of audacity." -- Benjamin Disraeli
"Audacity, more audacity, always audacity." -- Georges Jacques Danton
"The first quality that is needed is audacity." -- Winston Churchill