17 December 2009

On Width

Update: a different reliable source indicates the land impacts along the corridor were determined (and minimized) by laying out an 87-foot wide corridor, measured over the fence footings on either side. That was a nominal value, with narrower design exceptions possible in highly-constrained locations.

Original Post: When we reviewed just how wide the Caltrain corridor is, we assumed a 75-foot minimum width for a four-track rail corridor, measuring from fence to fence. New figures, obtained from reliable sources, indicate typical dimensions will be quite a bit more:
  • 15' (4.6 m) Caltrain track spacing (measured center-to-center)
  • 16'6" (5 m) HSR track spacing, or between HSR and Caltrain tracks
  • The kicker: 23'6" to 28'6" between the outer track center line and the boundary fence, to allow for third-party utility easements (as already exist in many places along the Caltrain corridor), overhead electrification poles, maintenance walkways, drainage structures, etc.
Worst case, that adds up to 28'6" + 16'6" + 16'6" + 16'6" + 28'6" = 106'6" (32.5 m). The best case, with Caltrain tracks in the middle, adds up to 23'6" + 16'6" + 15' + 16'6" + 23'6" = 95' (29 m). Both figures are measured fence-to-fence, presumably for a situation where all four tracks run at ground level.

Time to Panic?

Well, maybe not quite yet.

These figures are quite likely quoted for the nominal situation, where plenty of land is available. Indeed, more than two-thirds of the peninsula rail corridor is 100 feet or wider, allowing generous side clearances. Caltrain's own environmental documents, drafted for the electrification project long before HSR came along, include the typical four-track section reproduced at right, with a nominal fence-to-fence width of 89 feet.

Then again, the HSR numbers are incredibly generous by international standards, and probably accommodate vehicular access along both sides of the right of way. (A Department of Homeland Security Crown Victoria is 6'6" wide, for reference.)

The minimum legal side clearances in California are dictated in CPUC General Order 26-D. The absolute minima are 14' between track centers and 10' side clearances to the edge of the right-of-way. Caltrain's engineering standards reflect these constraints in a clearance drawing. None of these standards envision trains running at 125 mph and above, since those have never existed in California. Better numbers can be gleaned from foreign standards, for example the German Eisenbahn, Bau- und Betriebsordnung (EBO). That particular standard requires the following side clearances:
  • A danger zone (free of any obstructions such as poles, walls, etc.) of 2.5 m (8'2"), measured from the track center line, for train speeds less than 160 km/h (100 mph) and 3.0 m (10') for greater speeds.
  • Outside of the danger zone, a 0.8 m (2'6") space for rail personnel to take refuge at a safe distance from trains
  • The danger zones of neighboring tracks may overlap, with tracks spaced 4 m (13') for speeds less than 250 km/h (150 mph) and 4.2 m to 4.5 m (14' to 15') above. Since European trains are about a foot narrower than ours, that's roughly consistent with the 15-foot minimum already in use on the peninsula.
Where the Caltrain corridor is too narrow (such as in certain sections of San Mateo or Menlo Park), technical and political considerations make it probable that the HSR project will prefer adapting to the local constraints before seizing property and revving up the bulldozers.

If one had to make an educated guess about the minimum allowable corridor width, as opposed to the typical width, it would likely be closer to the values in the German EBO and/or the European Technical Specifications for Interoperability (TSI). Using the 16'6" track spacing mentioned above, that comes out to 0.8 m + 3 m + 5 m + 5 m + 5 m + 3 m + 0.8 m = 22.6 m, or 75 feet. The actual minimum for the peninsula corridor is likely to be spelled out in the upcoming draft Analysis of Alternatives.

Those figures are only valid for an alignment at grade level, which is the narrowest option. Raising or lowering the tracks requires additional width for construction.


  1. Bigger is better. And more concrete and more bulldozers always beat being smarter. Because we're Americans. Our feet are bigger and better than meters. And Jesus writes our engineering specifications.

    Our CBOSS will whup your pansy ETCS, just like our double-wide extra ass space super-sized tracks spank Euro-weenie one-track-and-one-platform-fits-two-kinds-of-trains.

    Just wait till you see our designed in California super-trains.

  2. You know, the Crown Vic won't be in service in 10 years. Production ends in 2012, we'll see specs on its replacement soon - it'll probably be narrower.

  3. Actually, I've seen both 16'6" and 15' numbers specified in CHSRA EIR documents. My guess is that for the slower (125 MPH) sections, 15' track spacing is sufficient.

  4. Does STRACNET have an width implication? Does adjacent freight operation have any width implication? Does proximity to structures/people have any width implication?

    And is it crown victoria's that move along access roads? Or is it service vehicles (some of pretty heavy type) and emergency vehicles like ambulances and fire engines? How wide are those? Wider than a crown victoria?

    In areas that are heavily neighborhooded - places where ROW is typically slimmer, are places where they'll be inclined to actually go to wider clearances - proximity to human life, proximity to structures like homes, schools, parks. Other than a few shortsighted/greedy CHSRA officials, I don't think they will feel much like skimping on safety clearances in these areas.

    But wonders never cease to spew forth from the office of the CHSRA - so we'll see.

  5. What sort of ignorant, unprofessional, unqualified, uneducated idiot even considers headspans for open track? On other continents, we've thrown lots of good money undoing the bad money "saved" by such cheap and unreliable construction of 50 to 100 years ago. (For example this in Britain isn't a headspan, but it's just as bad and just as obsolete and just as expensive to have to come and and do right the second time over.)

    What sort of moron writes engineering specifications for what is supposed to be modern system in inches? Why not pick your own track gauge while you're at it?

    Worst of all, what sort of terrible management allows this sort of special snowflake, we'll show the rest of the world, fuck you all, we know better, CBOSS-level insanity to go ahead? It's one bad thing to have fifth-rate, know-nothing, unqualified, unemployable except in America engineers with wretchedly bad judgement and complete ignorance of the last 100 years of rail development that doesn't involve moving coal or petrochemicals, it's quite a bigger and more catastrophic management and public policy failure to let a bunch of chimps piss away several billion public dollars ... money that could be spent on hospitals, or on a good rail system, or to pay school or university teachers, anything but this.

    I'm an engineering project manager in the private sector, and I'd be fired in about two seconds I even considered this level of make-work, Not Invented Here, failure is the only option square wheel re-invention.

    Who exactly supposed to be in charge of CHSR and Caltrain's quite obviously grade school level engineering teams? Is there anybody anywhere who has any financial and professional skin in the game or is it just the public who are going to be shafted?

    And lastly, where do they even find the quality of person to do this? Where I live you need years of education to learn to be qualified to be a rail engineering professional. Much of this is understanding engineering best practice, why engineering standards exist, and how it is very much more important to be competent and to be reliable and to mesh with existing practice than it is to run amok and build your own stupid expensive isolated little playpen of failure.

    Where on earth do they find these people? Shouldn't you have laws against having inmates run your asylums?

  6. Realistically, CHSRA is going to have to get creative about how to make do with the very limited space available. Standard US thinking regarding "bigger is better" doesn't make sense in the Caltrain corridor.

    The 16'6" track separation looks very generous to me. Lots of countries operate trains at relative speeds of 250mph with less than that.

    As for the access path, 23'6" to 28'6" is measures to the track centerline, so deduct 5-6 feet for the actual train envelope and another 8-9 feet for a safety envelope. Shaving some of that width might require switching to narrower special vehicles for maintenance crews not just for the railway but also the utility companies. Buying a fleet of those would cost a lot less than adding even just one foot of width to the right of way and building wider grade separation structures.

    Similarly, if CPUC has a 14' minimum, why pad it to 15' to begin with? It's not as if 14' were narrow by international standards.

  7. Wrt to the narrowest sections of the corridor: the political issues may indeed force track stacking. Full grade separation there would imply one of three options:

    - Caltrain/UPRR at grade, HSR elevated, deep road underpass (problematic where there is a major frontage road but no change in access to freight spurs)

    - Caltrain/UPRR at grade, HSR underground, tall road overpasses (ditto plus the whole trench/tunnel can of worms)

    - Caltrain/UPRR elevated, HSR underground, roads at grade (worst of both worlds for rail, most expensive, but maximum flexibility for cross road patterns. Any unused space between grade level and elevated tracks could be developed for commercial use, but this would need to be integral to the design from the outset as it involves additional measures to keep noise/vibration there tolerable. This does not apply for vehicle parking.)

    Funky bonus variation: Caltrain+UPRR+HSR northbound elevated, HSR southbound in trench/tunnel.

    If none of these options are selected, any remaining grade crossings for Caltrain will make it difficult to raise service frequency to the target of 10tph for the commuter line alone. At an average 90 seconds per closure event, the crossing gates would be down 50% of the time during rush hour.

    While FRA mandates full grade separation for HSR on safety grounds, Caltrain really needs it to avoid snarling up the cross roads for decades to come. Stacking with any tracks at grade would effectively reduce the number of options for grade separation later on to just one and even that only if other, non-rail factors permit it.

    Better to bite the bullet and widen the right of way to the 75 feet minimum indicated in the post.

  8. One factor in width considerations is the impact of shoofly tracks during construction. Where the ROW is already wide enough to accommodate two of them, they are arguably the best way to keep Caltrain/UPRR operating while four grade-separated tracks are constructed to one side of them.

    Where that width is not available and widening beyond 75' politically infeasible even if ED takings are extra-generous, the following options may be available for construction above ground:

    a) build just a single shoofly track

    b) move legacy tracks laterally one at a time, build two grade separated tracks next to them, switch the legacy services to these new tracks, build the other two. Retained fill embankments would feature twice as much concrete for this case. Aerials would have to be split, precluding switches between the inner tracks (or alternatively, switches between the inner and outer tracks). If all four tracks remain at grade because the locally preferred alternative is deep under- or tall overpasses, such switches could be added at a later date.

    If the ultimate track order is SFFS or FSSF, one track would be used in the "wrong" direction during construction, with consequences for signaling. The strategy might also require adding switches that will rarely be used once the line is fully built out. This does not apply for the track order options FFSS and SSFF.

    Special measures might be needed to adequately protect construction crews and facilitate logistics. Also, there would be precious little room for storing bulk materials on site, instead construction crews would have to implement just-in-time principles.

    UPRR might be limited to a single track during construction in any sections built two tracks at a time. However, that may not be necessary if slab track and suitable slow orders are used to prevent the freight trains from causing excessive wear and tear on the brand-new rails that would later be dedicated to HSR..

  9. @ anon @ 11:14 -

    afaik, the SF peninsula is still part of STRACNET.

    However, I doubt the designation will have much impact on the design, as long as Caltrain/UPRR remain above ground.

    The AAR loading gauge diagram is a little hard to read, both because it includes everything but the kitchen sink and because of the poor scan quality. The STRACNET gauge is indicated as "DOD outline".

  10. @ Rafael

    The AAR loading gauge diagram is easy to read if you remember to view it in its native/original size. Many web browsers default to "scale-to-fit" when displaying JPGs and GIFs.

  11. Morons.

    Utter, boundlessly ignorant, limitlessly incompetent, inconceivably unprofessional, morons.

    BTW here's your CONTEXT SENSITIVE way to run four 200kmh tracks.

    THROUGH OPEN LANDS, on new inter-city construction, ***NOT*** THROUGH the middle of suburbs.

    Check it out.

    4.7m+*12.1*m+4.7m pair track spacing.

    6.5m worth of service roads on BOTH sides of the tracks, starting 7.95m from the outermost track centreline.

    Noise walls.

    Cable runs.

    No INSANE, 50 year obsolete, maintenance nightmare, reliability nightmare, flexibility nightmare headspans. I mean, who would be stupid and ignorant enough to even propose such a thing, but still have enough hand-eye coordination to be able to run AutoCAD to produce a cross section?

    A grand studly total of 49.75m (163'3") of HE-MANLY, over-the-top edge to edge raw pulsing trainly girth. Caltrain's highly skilled engineering division doesn't even hold a candle at a mere 95 feet. (And add another 2m (track to pole clearance) total to the OeBB section if track speeds are over 200kmh.)

    But ...

    But ...

    But ... nobody in Austria would be as bat shit insane as to propose such in a suburban area.

    But ... on the other hand nobody in Austria (or Germany, or Sweden, or France, or Switzerland, or Italy, or Spain, or ...) would be as bat shit insane as to even dream they were going to run at 200+kmh through century old suburbs, 350kmh right though the middle of regional city centres, unnecessarily track-share a suburban access corridor for 80km, propose to operate incompatible trains with incompatible platform heights along a very constricted corridor, propose two separate urban terminals in one city for two gratuitously incompatible train types, exclude the 75% of daily bread and butter residents from even reaching the major CBD station, propose their own Looney Tunes signal system, or claim with a straight face that any of this has a snowballs chance of being delivered to any sort of schedule or budget, or that it won't simply be a catastrophe for potential regional transit all the way from SF to SJ.


    Out of control, unsupervised, kindergartner morons. All those in house designs for square wheels are going to work out real good, for sure.

    CBOSS seems to be just the $300 million tip of the iceberg of solid excrement.

  12. Just FYI here is a simple sketch, done at the appropriate crayon level, of the evil inappropriate NIH not special and unique enough Teutonic EBO track spacing law quoted in Clem's article.

    This sort of thing fits very nicely nearly everywhere along out sad and about to be ruined by monumental incompetence transportation corridor.

    You'd almost think that some other people somewhere else might have thought through what is and isn't needed or desirable or mandatory .. or insane.

  13. Not that caltrain doesn't deserve ridicule for some of their engineering decisions, but they're not proposing headspans for the entire route, they're proposing OCS poles, and only using headspans for the tight areas where they cant fit poles:

    The messenger wire is supported by means of cantilevered, hinged bracket arms that extend horizontally over the track from vertical steel poles mounted clear of the dynamic envelope of the vehicles. These poles must be placed within 10 to12 feet of the centerline of the tracks they serve. In complex areas, where there is limited clearance between tracks, multi-track support structures, such as multi-wire headspans attached to taller steel poles, are employed.

  14. For what it is worth, since I wsa not there, I was told the CalTrain at technical working groups stated 85 feet is what they needed.

    I believe that was to construct above grade tracks on a concrete wall. Berm type of construction, such as they used in San Carlos needs much more, perhaps as much of 130 feet for the 4 tracks.

    Downtown MP has 55 feet is some spots, and a lot of 75 feet along the corridor.

  15. @Matthew, I was of course kidding about the Crown Vic.

    @Rafael-- narrower special vehicles? Have you ever heard of a hi-rail vehicle?

    @Richard & anon... I know this drives you nuts, but please go easy on the ad hominem attacks. It may feel good to unload but I doubt it's very effective. In fact it may be counter-productive: it discredits the many good ideas you have. In a nutshell: rail at designs, not people.

  16. Downtown MP has 55 feet is some spots, and a lot of 75 feet along the corridor.

    For a whole block if I'm reading the maps correctly. Bordered by parking lots on the west side. The side were they would have to take land to widen it out. How vital are a block of parking lots to the historic and cultural milieu of Menlo Park

  17. @ Clem -

    actually, I hadn't heard of hi-rail vehicles before. Using those could eliminate the need for access roads altogether, in those sections where there is currently not enough room for them.

    However, there is a price to pay: the track in question would have to be closed to rail vehicles at that time. Not a problem if even minor inspections and maintenance in the affected sections of the Caltrain corridor are done exclusively at night, IDK if that's an acceptable restriction given the proximity of residential homes. This restriction would extend to the utility company as well.

    Looks like the issue of width won't be resolved without some planning effort regarding off-design operations and maintenance procedures.

  18. Clem, Sorry, I don't understand your Update.

    When you say land impacts were determined, are you saying that land impacts (ie: eminent domain needs) were determined using a corridor width that is too small to be realistic? (8 feet narrower than your 'best case' number).

    Or are you saying that your second source says they are already using 87 feet as their width requirement so your first sources is wrong?

    And when you say 'were determined' does that mean in the first program EIR, or is tha what they are using now for the updated EIR?

  19. However, there is a price to pay: the track in question would have to be closed to rail vehicles at that time.

    One of the features of ERTMS ( and ACSES ) is that it can close down track when there are people on the track. Since it can do that I suspect it was designed in because they expected that to happen frequently. Almost all of the maintenance workers and all of the people in emergency services have a skill called "walking" too. They can "walk" from the places where the ROW is wide enough for the service vehicles the half block or so to the place where they can't drive the service vehicle.

  20. @ adirondacker12800 -

    a short walk would be fine for inspection tasks. It's when maintenance workers have to bring in a lot of tools or materials that a vehicle is preferable.

    If CHSRA, Caltrain and the utility company can figure out maintenance procedures that don't need an access road in the sections that are too narrow to accommodate one, more power to them.

  21. @anon, I think the way they are going about it for the project EIR is (was--it's already done) to lay out an 85-foot corridor (+2 feet for fence footings) all the way up the peninsula, and jiggle it around until it fit as best as possible within the existing railroad land. Then they looked where it didn't fit, and decided where exceptions would be needed (narrower than 85 feet) versus where acquisitions would be needed, presumably as a last resort.

    As I wrote in the post, I suspect the original numbers I quoted are for a typical, unconstrained case where plenty of room exists. They would not use those numbers in "tight" spots.

    The two sources are not necessarily contradictory. There are different widths being discussed, for example 75 feet exceptional minimum (requires design exception), 85 feet nominal minimum (without design exception), 95 to 105 feet desirable range. When you hear a width number, it's important to know what's what.

    Hopefully the draft AA will shed light on this issue.

  22. On european width: the difference ain't that big. European cars must fit to loading gauge 3150 mm wide in curve with radius of 250 m. So naturally if the car is longer, it gets also narrower. For example, Talgo 350 at 2960 mm is a bit narrower than american BiLevel coach, but certainly not even close to a full foot (305 mm) difference. Some european trains are even wider like Bombardier Spacium (3060 mm) or two-axled railbuses (3120 mm).

    Regarding track spacing: the 0.8 safe zone must be next to every track. Common design is to put these zones between outer and inner tracks. Apart from being safer for track workers, this design also allows more effective OCS support placement.

    CHSRA and Caltrain should IMO evaluate the option to switch to wider car bodies. Width around 3300-3400 mm allows for comfortable 3+2 seating arrangement providing the same capacity per lenth of train as double-deck cars with 2+2 seating.

  23. Adirondacker1280020 December, 2009 11:20

    It's when maintenance workers have to bring in a lot of tools or materials that a vehicle is preferable

    The infrastructure they will be maintaining carries vehicles. Vehicles that are capable of carrying people, their tools and material. They can use one of those vehicles. A solution so common that even the model railroaders can get MOW cars in a wide variety.