26 June 2010

Tunnel Sizes

Mountain View mayor Ronit Bryant argues that a deep tunnel for Mountain View ought to be considered because "most of the BART in our area is underground. If they managed it then, why can't they do it now?" Here's one big reason why:

The HSR tunnel requires large clearances around the train for aerodynamic reasons. Without those clearances, a high-speed train (going 1.5 to 2 times as fast as BART's maximum speed) will dump megawatts of power into swirling the air inside the tunnel, making it unbearably hot. In the diagram, the BART tunnel bore is 16 feet (5 meters), and the HSR tunnel bore is 26 feet (8 meters).


  1. WOW!!! They have BART in Mountain View??? When did this happen? Does BART know about this?

    And were these new BART tunnels in Mountain View dug in the 70's with the original system, or were they dug recently, under modern digging regulations?

    ...or is this just smoke being blown?

  2. I suspect it is, but if you're going to blow smoke, might as well blow informed smoke.

  3. Not to mention the necessary clearance from the tunnel walls for the high-voltage (25kV) overhead catenary vs. the third rail (1kV) of BART.

  4. ""most of the BART in our area is underground.""?????

    A large majority of the BART system is at or above ground. Or are Richmond, Fremont, Dublin and Pittsburg figments of my imagination?

    (An interesting exercise would be to work out if a majority of BART passenger-km were underground, given the relative efficiency of core SF-Oakland-Berkeley ridership, and the ghost trains that run to the exububs. The Quentin Kopp Memorial PB-Tudor-Saliba-Soprano cut and cover extravaganza Colma-Millbrae could potentially screw up the numbers.)

  5. That article's an interesting read for many reasons. The council's apparent reaction to an honest resident's thoughtful presentation speaks volumes.

    Presenting a picture of an aerial train platform that runs through Berlin,
    Germany, resident Jarrett Mullin urged the council to reconsider the affordable
    aerial platform option, which was unpopular at a recent community hearing on
    high speed rail. He said the space underneath provided many opportunities for
    pedestrian and transit connections, and also storefronts.

    I've been to Germany and that elevated train dominated everything around it,"
    said council member Jac Siegel. "It's not very nice. You don't want to be around
    it very much."

  6. It only "dominates everything around it" if there's nothing around it. The council member is an idiot.

  7. The initial BART system breaks down roughly as follows. (Mileage taken from Google Earth and rounded. Those wishing more accurate mileage may measure it themselves. I realize my rounding doesn't add up to current totals.)

    Oakland Wye -Fremont:
    Tunnel- 1 mile
    Not Tunnel - 23 miles

    Oakland Wye - Concord:
    Tunnel- 4 miles (3 under Berkeley Hills, which would require a tunnel anyway.)
    Not Tunnel - 17 miles

    Pittsburg/BayPoint Extension:
    Not Tunnel - 6 miles

    MacArthur - Richmond:
    Tunnel - 1 mile*
    Not Tunnel - 9 miles*
    * Berkeley paid for two additional miles of tunnel from a city-wide property tax assessment, so what was built was three miles of Tunnel and seven miles of Not Tunnel

    Oakland Wye - Daly City:
    Tunnel - 5 miles
    Not Tunnel - 4 miles
    Transbay Tube - 5 miles**
    ** Funded as a bay crossing from Bay Bridge tolls

    Daly City - SFO & Millbrae:
    Tunnel - 6 miles
    Not Tunnel - 4 miles

    Bayfair - Dublin/Pleasanton
    Not Tunnel - 13 miles

    So, Original System:
    Tunnel - 13 miles
    Not Tunnel - 51 miles
    Transbay Tube - 5 miles

    Current System:
    Tunnel - 19 miles
    Not Tunnel - 74 miles
    Transbay Tube - 5 miles

  8. Clem, I don't support tunnels for the suburbs. But is the 8 meter diameter needed for 125 or 110 mph speeds, or rather from 220 mph operations? Or is the clearance around the high-voltage wires the bigger issue at these speeds?

  9. I think it's for 125, since CHSRA has specified >30' bore diameters for 220mph

  10. @ Joseph Eisenberg -

    it might be possible to reduce the tunnel bore diameter at 125mph speeds by using an overhead conductor rail instead of a catenary wire. That solution is rated for commercial operations at up to 230km/h (~140mph).

    However, it's a moot point because a bored tunnel should not be needed in Mtn View. If CHSRA insists on quad tracking the section between 85 and 237, it may be preferable to replace three overpasses with designs that do not require supports within the transportation corridor of Central + rail + W Evelyn.

    Cities can ask for the moon. That doesn't mean they will or should get it.

  11. Rafael wrote: "it might be possible to reduce the tunnel bore diameter at 125mph speeds by using an overhead conductor rail instead of a catenary wire."

    This issue is cross-section (clear air space, aerodynamics), not vertical clearance.

    Fewer coughed up the buzzwords and more thinking, please!

  12. I wonder just how many different anonymous trolls there actually are on the transit blogs. Clem, are the IPs all from one person, or are there multiple anonymous snipers?

  13. Re posters' IP addresses: anybody can use a redirector or VPN elsewhere or grab the neighbour's open Wifi or connect from a cafe. (I for one have done all of the above.) It doesn't tell you much if you're trying to catch somebody who doesn't want to be caught.

    For the record and to establish my authenticity, "the Peninsula Rail Program are unprofessional morons of no technical ability."

  14. Adirondacker1280029 June, 2010 20:57

    Re posters' IP addresses: anybody can use a redirector or VPN elsewhere or grab the neighbour's open Wifi or connect from a cafe.

    You are assuming the average troll knows what an IP address is. The trolls here are average. I'd be more specific but that would be off topic and might give the trolls ideas.

  15. A costly CHSR requirement is the need for grade separations in urban areas and tunnels through mountains and below San Francisco Bay. The present 21.3 foot CHSR minimum height requirement would require 2.9 times the material to be removed for circular tunnel construction compared to the quantity of material removed in order to accommodate the 12.5 foot minimum clearance above the rails stipulated in a CPUC waiver granted during November 1964 at BART’s request. CHSR urban third rail electrification could reduce the CHSR urban tunnel height requirements for single level rolling stock to a figure approaching 12.5 feet. Note that all 25 KVAC powered rolling stock drop traction power voltage sharply within locomotives or MU’s before applying current to the traction motors; often to a level approaching third rail compatible voltages – 1,500 VDC within TGV locomotives for example. An “or” gate circuit topology within a CHSR dual voltage powered vehicle would prevent a rolling stock induced short between the 25 KVAC and third rail 1,500 VDC power supplies. Peninsula freight traffic north of Lawrence should be diverted over the Dumbarton rail Bridge and the center of the 101 Freeway to South San Francisco. The proposed Caltrain electric trains weigh 44% more per seat than BART cars. Double deck cars should have a lower weight per seat but that weight advantage is clearly overwhelmed by required freight train territory frame strengthening and traction power transformers. Note that maximum transformer power density declines at lower power levels. Peak transformer power capacity is proportional to the magnetic field cross-sectional area times the current path cross-sectional area but the weight of any object is proportional to the cube of any linear dimension. The transformer weight and overhead clearance penalties to accommodate 25 KVAC power distribution become particularly onerous for urban local rail service.

  16. BART has a ridiculously small loading gauge, not really suitable for high speed trains (even without catenary), and definitely too small for Caltrain bilevels (with or without overhead wires). Plus, you have to have considerable aerodynamic clearance anyway, even at 125mph.

  17. @ anon @ Jun 29 10:08 -

    The way they deal with the aerodynamic envelope in the Channel Tunnel is with pressure equalization ducts connecting the single bore tunnels every 250m. These are in addition to connections to the separate service bore.


    The bores in that tunnel are 7.6m (25ft) diameter, suitable for the Eurotunnel car/truck shuttle trains, which are extra-wide and extra-tall. This refers to the inside diameter, the TBM diameter was actually 8.6m.

    I didn't mean to imply that using overhead conductor rails could reduce those numbers by a large amount, more like a couple of feet.

  18. Adirondacker1280030 June, 2010 10:42

    The way they deal with the aerodynamic envelope in the Channel Tunnel is with pressure equalization ducts connecting the single bore tunnels every 250m.

    And they still have to have forced cooling. If I remember correctly, without the cooling the tunnel would reach 140F.

  19. The going rate for single track 200kmh running (and yes, 200kmh SJ-SF is all of insane, unjustified, and unjustifiable) is a bit under 50m^2 of free air space. (And yes, smaller has been done and is being done, but that's representative enough.) This means ~9m inside diameter. This means that squeezing every mm out of the system height of the overhead contact is not of practical consequence. (And yes, conductor rail systems have other advantages.) Not only that, but low-profile overhead can still be catenary overhead.

  20. And sadly, evacuated-air tunnel technology is not yet ready for prime time...

  21. @Rafael: note Channel Tunnel speed limit is 100 mph, less than what is planned here on the peninsula. Hence the similar / larger tunnel cross section.

    @John Bacon: the issue is not the size of the train. The issue is the size of the air cushion between the train and the tunnel wall. That free space needs to be very ample in order to allow 125 mph operation without overheating the tunnel. This has little or nothing to do with train design.

  22. @ adirondacker12800 -

    If the Channel Tunnel is indeed forcibly cooled, that may have more to do with the temperature of the surrounding rock than with aerodynamic heating. The deeper you go, the hotter the rock strata get. It's just a consequence of natural radioactive decay.

    Also, since it's a subsea tunnel, it slopes upward on both ends and there's virtually no elevation difference between the portals. Ergo, there's no natural convection.

    In California, none of the tunnels are likely to suffer from this problem. The surrounding rocks will heat the mountain tunnels a little, but they'll be on 3.5% inclines, so there will be a mild natural chimney effect. Not sure if it'll be enough to keep ambient temps low enough. Any aerodynamic heating comes on top of that.

    For tunnels in urban/suburban areas, rock heating should be a minor if not negligible effect.

  23. Adirondacker1280001 July, 2010 08:24

    If the Channel Tunnel is indeed forcibly cooled, that may have more to do with the temperature of the surrounding rock than with aerodynamic heating.

    It isn't a diamond mine. It's 75 meters or 246 feet at it's deepest. Took ten seconds to find the Wikipedia article that tidbit comes from. 250 feet under sea level, under the sea itself, without checking references is going to be the annual average temperature of the sea give or take a degree or two.

    Also "The Channel Tunnel was the first mainline railway tunnel to have special cooling equipment. Heat is generated from traction equipment and drag. The design limit was set at 30 °C (86 °F), using a mechanical cooling system with refrigeration plants on both the English and French sides that run chilled water circulating in pipes within the tunnel"
    with a link to footnotes.

  24. Palo Alto is currently the community most strongly advocating tunnel grade separations. When they contemplate the minimum disruption and moderate cost overhead bridges at the North Alma Street and Charleston Street crossings plus a Churchill Avenue pedestrian and bicycle only tunnel beneath Alma with the railway left at grade compared to the extraordinary cost and heavy construction activity required for a deep tunnel approach the desire for this expensive solution is likely to evaporate.
    But consider the grade separation problem near the Menlo Park station. A Menlo Park Transportation Department grade separation study came up with four different approaches all requiring the destruction of all nearby trees, a significantly wider property taking than the final result would occupy, and either a permanent 30 foot continuous wall through the center of town or a very expensive deep open cut. Accommodating a 25 foot above the rail loading gauge left no practical alternative. If the vertical loading gauge was reduced to 14 feet by using low overhead clearance requirement rolling stock and electrification structure a grade separation effort would have the following properties. The completed HSR and Caltrain shallow trench grade separation’s overhead crossing bridge pier and retaining wall concrete structures would be to the side of the present track structure. Therefore a future grade separation’s heavy concrete structures could be constructed in slit trenches on both sides of the present track structure without moving the present track or stopping rail service. Those structures could be lighter and yet be initially stronger than is usual for new concrete structures because they could then be backfilled and soaked in water for an extended period in order to let the concrete set while rail service can resume at normal speeds. For long sections such as for example the four crossovers near downtown Menlo Park a long summer weekend could be reserved when rain interference and high rail traffic demand is unlikely. Fill and tracks could be removed and reconstructed in the resulting trench. The enormous amount of fill to be removed could be transported in hopper cars. Through weekend rail service could be connected with a bus bridge. Because the below grade trench surface would be largely undisturbed and packed down for thousands of years a stable replacement track could be quickly constructed. The resulting moderately below grade station could be readily integrated with at grade walkways and building entrances. Entrepreneurs could easily develop attractive properties adjacent to or above the station.

  25. CalTrain Bilevels are already 16' tall (above top of rail). The large vertical clearances required for trains are why it makes more sense to put the road under the rails, rather than the other way around.

  26. Ambient air temperature rise within HSR tunnels is a function of air friction times the distance from the last vent. It is quite understandable the English Channel Tunnel required expensive refrigerated cooling given that no mid channel vent could sensibly be constructed along the 22 mile under-sea stretch between England and France. But any Peninsula right-of-way grade-separation tunnel vent frequency would be enough to prevent tunnel air heating from having a material effect on tunnel air temperature.
    However if the line were extended to the 3.6 mile BART tunnel under the Bay a maximum credible traffic density would raise its ambient air temperature significantly. Seven hundred foot trains on a 72 second headway would generate a 4.4 degree C temperature rise at 80 mph. If both Market Street Subway levels were connected to the Trans Bay Tube a 36 second TBT headway would generate an 8.8 degree temperature rise. (Note: The present BART train separation safety standard is H = v/2b + L/v + 1 seconds with a 3 ft/sec safety braking rate in tunnels and L = train length plus minimum train separation equal to 864 feet. Moving block position detection and platform edge access control screens could limit close-up and dwell times to 36 seconds each at Montgomery and Embarcadero Stations.-Longer than 700 foot CHSR trains could be assembled from CHSR limited and express peninsula sections and CHSR East Bay express sections converging in San Jose.)