05 July 2009

Focus on: San Mateo

The city of San Mateo was incorporated in 1894, 30 years after the railroad began operating. Along the 4.5 miles of rail through the city, there are today three Caltrain stations, two of which (Hillsdale and San Mateo) are Baby Bullet stops. The three stations in San Mateo generate greater combined Caltrain ridership than San Jose, a city with ten times San Mateo's population.

San Mateo is among the most treacherous areas on the peninsula to expand to four tracks for high speed rail because the existing Caltrain tracks run through dense residential neighborhoods and downtown blocks where the available right of way width is significantly less than 100 feet. The city is a "ground zero" of future grade separations: it is home to one-fifth of the grade crossings on the entire peninsula rail corridor (9 out of 46). To improve safety and traffic circulation, the city has long had plans to grade-separate all of its rail crossings; high speed rail will only hasten this process. The San Mateo County Transportation Authority has been carrying out a series of preliminary grade separation footprint studies in support of these plans.

San Mateo finalized a Rail Corridor Transit-Oriented Development Plan in April 2008, covering the vicinity of the Hillsdale and Hayward Park Caltrain stations and including the re-development of the Bay Meadows racetrack site.

Station Area and Downtown

The San Mateo Caltrain station was rebuilt in 1999-2000 to alleviate auto traffic congestion. The new $11 million station with underground parking (photo above by ibison4) replaced a sparse older station a few blocks to the south. Every time a train stopped at the old station's center boarding platform between 3rd and 4th Avenues, auto traffic on 2nd through 5th Avenues was paralyzed, exacerbating rush hour congestion.

With some foresight, albeit a bit fuzzy, the new station's underground garage and surrounding structures were reportedly built so as to allow the tracks to be relocated underground. This would not only grade-separate the tracks through town, but also accommodate a future BART subway, as was envisioned in the late 1990s while construction of the SFO extension was underway.

Plans change.

The problem is now this: the new station building and underground parking encroach on the railroad right of way and do not leave the necessary clearance for four tracks, as likely required for Caltrain and HSR, whether at, above, or below grade. Worse, a couple of blocks south of the station, San Mateo's new downtown cinema and Main Street parking garage were completed in 2003 and encroach on what was once railroad right of way, leaving just 50 - 60 feet for the tracks (see Caltrain right of way map).

There will be no easy or cheap solutions to these constraints. In discussions about a preliminary grade separation footprint study for downtown, city staff makes the following key observations:
  • The vertical alignment of all the downtown rail crossings must be consistent, since their close spacing does not allow sufficient space for rail grade changes;
  • Any grade separation alternatives with the tracks at grade or near grade is infeasible because of the need to lower or raise downtown streets in a manner that severely impacts frontage on both sides of these streets;
  • The two remaining vertical alignment alternatives leave the streets at grade, with the rails either fully elevated or underground--the latter being the city's stated preference;
  • The right of way required to build a four-track underground trench is 110 feet wide, far more than the 50 - 60 ft available;
  • Moving the horizontal alignment of the tracks to the east (with impact to Railroad Ave and frontage) is preferable to impacting the new train station, Main Street parking garage and downtown cinema;
  • The underground alignment presents significant technical complications, with residential property impacts just north of downtown where the tracks would ramp down, possible street closures, as well as the technical difficulty of crossing San Mateo Creek underground.
While no decisions have yet been made, and despite the city's stated preference for a depressed alignment, the elevated alignment may present the least physical impact to property adjoining the tracks, although the visual impacts are worst. The city is considering options to mitigate the Berlin Wall effect, including building the track on a viaduct instead of retained fill. Contrary to San Mateo's official preference, the California High Speed Rail Authority shows an elevated track profile with 15-foot embankments through San Mateo in their program EIR/EIS; this may or may not be the preferred alternative once they complete the project EIR/EIS.

The issue of track alignment through downtown San Mateo is discussed in much greater detail in Threading the San Mateo Narrows. That discussion includes diagrams showing various track configuration options such as the elevated shown below.
For more details, read all about Threading the San Mateo Narrows.

North Central San Mateo

North of downtown, the tracks curve through dense residential blocks on right of way that ranges from 60 to 90 feet wide, less than the ~75 feet required to run four tracks on a retained embankment, and certainly less than the ~110 feet required to excavate a trench or tunnel. The tracks cross four residential streets (Poplar, Santa Inez, Monte Diablo and Tilton avenues) on low-clearance bridges that have long been slated for refurbishment.

The key question for downtown and the North Central neighborhood will concern the vertical alignment of the tracks. The close proximity of these residential blocks to downtown, combined with the gentle track gradients required for freight trains (yes, freight trains), will require that these areas be considered as one.

Vertical Alignment Considerations

The vertical profile of the existing tracks in San Mateo is shown in the figure below. This figure was created from Caltrain track survey data, with the vertical scale greatly exaggerated. The grade level of cross streets is estimated from the known vertical clearance at each location; the level of creeks is likewise estimated.

We make a few assumptions (for more background on where these came from, read about The Shape of Palo Alto):
  • Vertical track radius is constrained to a minimum of 10 km
  • Gradient limit of 1 to 1.5% for freight trains, 2.5% for passenger trains
  • 20-foot clearance from underpass road surface to top-of-rail (to clear trucks)
  • 30-foot clearance from overpass road surface to top-of-rail (to clear freight trains)
  • Tracks must be close to level for 750 feet north of First Ave, for Caltrain platforms
  • Tracks are elevated at Peninsula Ave., an area with adequate clearances and commercial frontage that would favor this minimum-cost solution.
Scenario #1: track elevated throughout. With a 1% gradient limit and a 10 kilometer vertical radius constraint, this is what San Mateo would look like. We assume each road is depressed by 5 feet (something that can be done with little impact to adjoining frontage), with the rails raised by 15 feet. It is likely that a viaduct structure would replace the existing embankments, possibly making room for a rebuilt Railroad Ave. All vehicle height restrictions in North Central would be removed. Worth noting, the elevated alternative does not require the closure of any crossings, unlike depressed alternatives.

Scenario #2: trench through downtown. To shorten the sloping approaches and reduce the depth of the trench, we assume each road is raised by 5 feet (something that can be done with little impact to adjoining frontage), with rails depressed by 25 feet. Even this 25-foot trench penetrates below the water table and will require constant pumping to keep dry. With the 1% gradient limit desirable for freight trains, this design would require the closure of Villa Terrace, Bellevue, Poplar, Santa Inez, Monte Diablo and Tilton avenues, unless those streets were either raised or depressed with considerable impact to frontage on either side of the tracks.

Scenario #3: trench through downtown, with steep approaches. With a less stringent gradient limit just a shade over 2%, difficult for freight trains but easily handled by high speed trains and Caltrain, the approaches to the trench are much shorter. This allows underpasses at Villa Terrace, Bellevue and Poplar and reduces the need to close cross-streets. Santa Inez, Monte Diablo and Tilton would likely still be closed.

Scenario #4: tunnels. While this alternative will no doubt be studied by the CHSRA, it is unlikely to be implemented due to a host of disadvantages, not the least of which is the astronomical cost. The sloping approaches to a tunnel diving under downtown are likely to present even greater interference with cross-streets than the depressed trenches described above.

Whatever alternative is ultimately favored, one must keep in mind that the biggest constraints on the vertical alignment of the tracks come from freight trains, which don't handle steep grades well and require very high vertical clearances. The more nimble electric trains used by HSR and Caltrain would easily handle steep gradients up to 2.5%, and would require about 3 feet less vertical clearance. In short, freight trains may have a direct impact on San Mateo neighborhoods.

Southern San Mateo

The southern half of San Mateo is an area currently slated for extensive re-development. It does not have nearly as much access across the tracks as northern San Mateo, and the available right of way is far less constrained, generally greater than 100 feet (see maps for mileposts 18, 19 and 20). The area encompassing the Hayward Park and Hillsdale Caltrain stations (and the former Bay Meadows race track) is slated for redevelopment as described in San Mateo's Rail Corridor Transit-Oriented Development Plan.

Hayward Park includes a reverse curve in the tracks (curving right, then left). The northern curve at milepost 18.8 has a radius of just 1100 meters, good for about 95 mph. This curve made #5 on our list of Top Ten Worst Curves on the peninsula. While it was already flattened by shifting the tracks 20 feet westwards in the year 2000, this curve may need to be straightened some more so that high speed trains don't need to slow down as much in this area. This may not please residents of South Boulevard.

The Hayward Park Caltrain station was rebuilt in 2000 and moved slightly to the south of its former location. The original station provided convenient pedestrian access across the tracks at 16th Ave; the City is now considering a new pedestrian underpass at this location to restore the access that was lost after the move. The new Hayward Park station was built for three tracks, although it currently has only two tracks. The southbound platform already includes a cut (see photo at right) allowing it to be trimmed back to make room for a third passing track. Whatever the original intent of this configuration, the station will likely be entirely rebuilt when high speed rail comes through town.

The Route 92 overpass (actually two side-by-side bridges, seen in the background of the photo) provides ample clearance for four tracks.

The Hillsdale Caltrain station is already planned for relocation to the site of the former Bay Meadows station, and will be located between newly constructed grade separations at 28th and 31st avenues, connecting the west side of El Camino to the newly developed Bay Meadows area. The tracks will be elevated over 25th, 28th and 31st, linking up to the existing grade separation at Hillsdale Ave. These long-standing plans by the Caltrain JPB and the City are unlikely to be altered by the HSR project, except for quadruple tracking; for a detailed description, refer to Chapter 4 of San Mateo's Rail Corridor Plan. The resulting vertical profile is shown in the diagrams above.

South of Hillsdale, the frontage along the east side of El Camino Real is already owned by Caltrain, Samtrans or Union Pacific; while businesses on that narrow strip of land will likely be displaced, plenty of land is available for expansion of the tracks.

San Mateo promises to be one of the biggest design challenges on the peninsula. The CHSRA certainly has its work cut out for it, with a 50-foot right of way threaded through dense city blocks surrounded with residential neighborhoods, with a creek thrown in for good measure.

NOTE: This post will be updated continuously, as warranted by additional information or new events relating to San Mateo.


  1. The Hayward Park station is actually big enough for four tracks with side platforms. The spacing of the existing tracks is rather wider than normal, so there's enough room for a fourth track in the middle, in addition to the track created by cutting back the southbound platform. It could have made for a reasonable mid-line passing section if the four track section started just south of downtown San Mateo and continued past Hillsdale (which would have two island platforms). Also, both Hillsdale and San Mateo are in fact Baby Bullet stops, each getting half of the peak-direction trains, with Hillsdale getting two of the five reverse-peak trains as well.

  2. If you were to cut back the southbound platform face, there would be about 46 feet between platform faces. That's not nearly enough space for four tracks.

    I have no idea why they built it that way... My best numerological guess is that somebody (in the incompetent 1990s era of Caltrain) had a spec with an 18-foot track-center requirement in stations, to provide clearance for a safety fence under CPUC GO-26. This being a future three-track situation, they couldn't decide how the spec applied, so they left open the option for possibly-required 18' centers by adding 3' to the southbound platform... concrete was poured before the spec was revised, and the rest is history.

    Thanks for pointing my error re: baby bullet stopping patterns. All fixed.

  3. If you were to cut back the southbound platform face, there would be about 46 feet between platform faces. That's not nearly enough space for four tracks....

    They are going to be rebuilding them for level boarding anyway, does it matter all that much where the ones they will be ripping out are located?

  4. You mention that the streets through downtown can be lowered by 5 feet with little impact on the existing buildings. But I'm just not seeing it. That would have a very significant impact on all the buildings along Railroad Avenue.

    Then there is the matter of expense. The split-grade solution through downtown would probably be much more expensive than just raising the viaduct an extra 5 feet, for questionable aesthetic benefit.

  5. Just playing devil's advocate here, but it seems to me there is an option #5: leaving the Caltrain/UPRR tracks at grade and running those for HSR on a viaduct featuring sound walls 25-30' above them (with/without AAR plate H clearance and overhead contact rails).

    Granted, even with an FRA quiet zone upgrade grade crossings are not desirable. Perhaps a couple of major crossings could be converted to deep underpasses for the railroad ROW plus the frontage road(s) in the context of such an option.

    I'm not saying that two narrow levels of tracks would be preferable to grade separating the whole lot, just that it's an option for San Mateo if/where eminent domain takings to widen the corridor are even less palatable.

    While it wouldn't be optimal, Caltrain could probably get by with a limit of 5-6 trains per hour by extending its platforms and operating really long trains.

  6. Rafael, why would your idea limit Caltrain to 6 tph?

  7. @Rafael: it is and has been San Mateo city policy to pursue grade separations. Your stacked concept is not a grade separation, and is really the worst of both worlds: an eyesore with little circulatory benefit.

    Besides, it's not really up to the CHSRA: the CPUC gets the final say on these things, being the regulatory authority on grade crossings. It is unlikely they would favor a non-separation.

    And nobody's going to short-change or hobble Caltrain if I have anything to say about it! Over my dead blog :)

  8. @ Alon Levy -

    Caltrain currently operated 5tph during peak periods. Full grade separation would help Caltrain to implement its 2025, which calls for growth to 10tph. If any grade crossings were retained, operating trains that frequently would have a significant impact on motor vehicle traffic across the railroad ROW.

    @ Clem -

    I'd much prefer to see the whole ROW grade separated as well, I was merely articulating an alternative. There's a trade-off between widening the ROW via eminent domain and retaining the downsides of grade crossings for Caltrain and UPRR only. It's about deciding which is the lesser of two evils. Personally, I agree with you that eminent domain will deliver the greatest long-term benefit to the community, at the expense of short-term pain. However, I don't live in San Mateo.

    Afaik, CPUC is only responsible for safety. While I'm sure they would strongly prefer full grade separation, they have allowed grade crossings for decades and would therefore have no basis for disallowing them for Caltrain and UPRR traffic.

    Btw: even if four tracks are elevated as planned, there could still be value in track stacking via a sufficiently tall viaduct rather than a retained embankment.

    First, it would permit diesel-based rail operations to continue during construction without any need for shoofly tracks.

    Second, the at-grade tracks could be retained for freight to give the viaduct designers more leeway on gradients and load limits.

    Third, iff freight trains are discontinued or can operate on the elevated tracks, the land underneath the viaduct could be put to other uses: parking lots, storage, shops and eateries, even a promenade ferrocarril.

  9. Two things: First, CPUC is responsible for safety, yes, but this includes grade crossing safety. If significant improvements are made to the rail alignment, they could say "no more grade crossings allowed" and that would be that. Second, building "stacked" tracks would almost certainly require shoofly tracks or periodic shutdowns of whatever is underneath. You don't want trains running under the stuff as you're building it, because it would be pretty bad if a chunk of concrete slipped off the crane and crushed a Caltrain.

  10. @ arcady -

    that's why you would move the prefab columns/slabs of concrete into position at night. How do you think they build new overpasses over existing freeways? Track stacking would indeed eliminate the need for shoofly tracks, at the expense of a taller viaduct.

    As for CPUC, those bureaucrats would get an earful from local governments if they insisted on eminent domain takings those communities had decided against.

  11. Clem, exactly how much width would be required for a four-track el? Would there be enough space for a four-across layout, or would CAHSRA have to stack the tracks two-over-two?

  12. @ Count Z -

    I believe the minimum track separation required by CPUC is 14 feet, so figure 56' as the absolute minimum for diesel. I believe Clem said Caltrain already has a couple of places with four tracks squeezed into just 60'. Allowing for OCS poles, the latter figure should be considered the rock bottom figure for the purposes of the Caltrain/HSR project. More width is required at stations, especially those at which HSR trains will stop as well. Note also that full-length platforms are 1320' (1/4 mile) long.

    For a variety of reasons incl. safe maintenance access and room for sound walls, fences etc. CHSRA is basing its design on a standard minimum width of 75' instead. It's not clear if it would consider making do with less in selected locations or, if CPUC would allow that given that HSR trains will be traveling at up to 125mph.

    2x2 track stacking would permit the construction of HSR tracks without widening the ROW anywhere, at the huge disadvantage of retaining grade crossings for Caltrain locals and UPRR trains. Moreover, as Clem reported in an earlier post, one of the reasons CHSRA chose the Caltrain corridor is because SP had had the foresight to purchase enough land for four tracks decades ago. Only in some locations did they later sell some of it off again to developers.

    The plan of record is to construct four tracks side by side throughout the peninsula, at whatever elevation is locally appropriate. The MOU with Caltrain leaves open the possibility of more than 4 permanent tracks, but that's just legalese for now.

    However, note that temporary shoofly tracks may be required during the construction phase. Where there is not enough ROW available for those, Caltrain may have to operate in streetcar mode on frontage roads, which would have to be restored after construction is completed.

    Track stacking during the construction phase only would avoid that, at the expense of taller aerial structures than would otherwise be required. Those would cost more to construct than retained fill embankments but the land underneath them would become available for other uses, including additional cross roads at grade.

  13. Clem didn't much mention the creeks flowing through downtown, or the 92 overcrossing. I wonder it they are non-issues? I would think that the elevated scenario must get pretty close to that overpass.

    But, as a neighbor to the tracks in North Central, none of this sounds good to me. Tunnels seem to be the best solution with the possibility of making use of the land on top for something. With 75-110 ft of ROW running from SF to San Jose it does seem like a chance to actually gain something, say a greenway for bikes, not to mention an opportunity for more prime business lots. Call me a dreamer...

    And one thing I haven't seen really usefully discussed yet (either here or elsewhere) is the noise of the trains? Here in San Mateo, as in other areas, the posh areas uphill don't currently get much noise from trains, but I would think that would change substantially with the elevation of the tracks and dramatically increased frequency of trains.

    Anyway, thanks for laying this out, Clem. I'll be sending the link to the neighbors here.

  14. 2x2 track stacking would permit the construction of HSR tracks without widening the ROW anywhere.

    Except for when they are digging the holes for the foundations. I suppose they could use piles or slurry walls to go all the way down to stable bedrock. Otherwise they are would have to dig enormous holes for the foundation work. It wouldn't be cheap. And in San Mateo we are talking about the station, a parking garage and multiplex that are less than ten years old, I doubt there is great historical significance attached to them. Much cheaper to tear them down.

    Clem didn't much mention the creeks flowing through downtown,

    An elevated structure avoids the creeks.

    If you followed the link to the post titled "The Joy of Tunnels" you'd come to this:

    "13. Building a 40-foot deep, water-tight barrier in an area with a high water table, extensive underground aquifers and surface creeks running perpendicular to the railroad is likely to cause changes in the water table, possibly leading to localized subsidence or flooding. Continuous pumping of surrounding aquifers may be required to prevent hydrostatic pressure buildup and excessive flotation (and cracking) of the sealed tunnel structure.

    or the 92 overcrossing. I wonder it they are non-issues? I would think that the elevated scenario must get pretty close to that overpass.

    The railroad would remain at the same grade it is now in any of the plans as it passes under 92.

    But, as a neighbor to the tracks in North Central, none of this sounds good to me. Tunnels seem to be the best solution

    Good! How are YOU going to pay for it? Increased property taxes? Increased sales taxes? Income taxes? Or a combination of the the three and maybe some others thrown in?

    Electric trains are much quieter than diesels, the grade crossings along with their noise, will be eliminated and thw whole Peninsula will be getting faster train service. Electric trains running on a viaduct or elevated or whatever you want to call it will be a net benefit to the Peninsula. If you want tunnels, I'm not going to pay for them. .... tolls on 101 and 280 in addition to the taxes! In perpetuity to pay off the construction and pay for the increased operating costa and maintenance.

  15. "that's why you would move the prefab columns/slabs of concrete into position at night. How do you think they build new overpasses over existing freeways?"

    Really? Funny. All that formwork I see must be safety netting?

  16. SP had had the foresight to purchase enough land for four tracks decades ago. Only in some locations did they later sell some of it off again to developers.

    @Rafael, SP had nothing to do with that. The new downtown station, the main street parking garage and the downtown cinema all went up in the last decade, well after it was known that San Mateo might some day require a four-track right of way. Somebody dropped the ball, big time.

    The cheapest solution may in fact be to bulldoze all three; but that's not a politically palatable solution, so look for impacts to occur on the other side of the tracks. Penny wise & pound foolish and all that.

    @Countz: roughly 70 feet is realistically the narrowest that a four-track elevated could get, but that doesn't take into account any land required for shoo-fly tracks during construction. I suppose that might not be much additional, if it's built in two 2-track halves.

    Clem didn't much mention the creeks flowing through downtown, or the 92 overcrossing.

    @Karen: are there any other creeks downtown besides the San Mateo creek? I wasn't aware; if there are, I'd like to know where they cross.

    As for the 92 overcrossing (grade separation!), it doesn't look like a problem to me. Refer to the vertical profile diagrams: 92 is way up top, near milepost 19.2

  17. @ Clem -

    "[...] if it's built in two 2-track halves."

    You mean if the aerial or embankment is built two tracks at a time?

    Also, could you elaborate on why you consider 70' rather than 60' to be the absolute minimum ROW width? I realize it "nice to have" additional room, but is there a hard regulatory or operational requirement for it?

  18. You mean if the aerial or embankment is built two tracks at a time?

    It can be, if space for a shoo-fly tracks is an issue. Otherwise, you need the entire width of the embankment plus another 30 feet for the shoo-fly tracks.

    Also, could you elaborate on why you consider 70' rather than 60' to be the absolute minimum ROW width?

    Because Spaethling and Litzinger said so.

  19. Clem-

    Even where there is sufficient ROW, though, there could be fun. Specifically, take the space to the other side of the San Mateo station. It looks to me (off the linked ROW map) that Railroad Ave is part of the ROW, but taking that space over would have impacts to the residents - there appear to be businesses and one residence that rely on that street as their only access...

    Note that this may not actually be sufficient ROW even including Railroad Ave. considering the space required for platforms etc., just using it as an example - how do we deal with sufficient ROW but that ROW would deprive access to businesses/residences.

  20. I haven't given the issue of horizontal alignment enough attention. I will try to fix that soon.

  21. Four tracks, with nastier and intrusive but space-saving portal overhead:
    0.5m foundation clearance
    + 3.8m side clearance (from catenary stanchions and fencing or sound walls)
    + 4.5m + 4.5m + 4.5m inter-track spacing
    + 3.8m
    + 0.5m
    = 22.1m ROW width (72.5 feet)
    > 51.5 feet = 15.7m (brilliant think-ahead planning minimum ROW width in lovely San Mateo = just enough for two tracks only.)

    This with no platforms or any station.

    Four through tracks with an island platform:
    - 4.5 + 1.6 + 9.0 + 1.6 (island)
    = 29.8m

    With stupid side platforms:
    + 0.7 (central inter-track fence)
    + 1.6 + 4.5 (platform edge and platform)
    - 0.5m (lose the outside fence)
    + 1.6 + 4.5 - 0.5
    = 34m

    Something has to give.

    Something to consider:
    + 3.8m
    + 4.5m inter-track northbound fast to north slow
    + 1.6m slow to platform
    + 9.0m platform width
    + 1.6m platform to southbound
    + 3.8m
    + 0.5m
    = 25.4m (= 83.3 feet)

    25.4 > 15.7 clearly, but 25.4 is a useful amount less than either 29.8 (FSPSF) or 34 (PSFFSP), especially in the context of the extraordinarily moronic things that the San Mateo City Fathers have done between First and Third avenues.

    Note also the 23.4m wide ROW north of downtown San Mateo, between Monte Diablo and Poplar. Squeezing four tracks (22.1m) in there is technically possible, but with no space to spare against the property lines, and with no scope for any curve improvement.

    Hmmmm... are four tracks operationally necessary at this point?

  22. Re: Hayward Park: close it.

    Once the new Hillsdale station is constructed north of the present location, Hayward Park will be almost within sight of that super-stop. (Hillsdale after all ought to be the centre of the Caltrain universe, served by more trains than any stations other than the terminii.)

    It's hard to justify Hayward Park's continued freeway-enjambed presence, and impossible to justify the expense and construction/operation compromise that will be required to rebuild it when Caltrain is all torn up for multi-tracking (with or without HSR.)
    Frankly there's much to be said for reopening Broadway than for spending millions to keep no-ridership white elephant Hayward Park on life support.

    San Mateo will still have two excellent-quality, closely-spaced, high-ridership stations.

  23. @ Richard Mlynarik -

    "Hmmmm... are four tracks operationally necessary at this point?"

    I'm not sure quite how to interpret your chart, but it seems to imply a cruise speed of 100mph for Caltrain - sounds a bit high. It also doesn't appear to account for the need to slow down in some of the sharper corners.

    Th fundamental principle of multiplexing Caltrain and HSR trains is interesting, though. It's perfectly reasonable to expect Caltrain will be the sole dispatcher on its own ROW, i.e. for HSR trains to be treated as guests in terms of operations.

    Perhaps you could articulate in more detail where/how HSR trains would pass by Caltrain locals. How about three tracks plus side platforms at e.g. the San Mateo station?

    I realize that CHSRA for some reason believes it must build a system that permits 12tph HSR trains, regardless of how Caltrain traffic grows.

    Your apparent idea of running only 4tph out of SFTT plus 6tph out of SJ Diridon isn't going to be popular with SF politicians, but just how much capacity is really needed out of that city? A full-length TGV Duplex (300km/h) has 1090 seats, the Japanese E4Max even has 1634 but it's limited to 240km/h.

    On a related point: AB3034 requires that it be possible to run non-stop express trains from SFTT to LAUS in 2h38m or so. It doesn't say the HSR operator has to have the freedom to schedule those during Caltrain's rush hour.

    Final note: in theory, you can build a station with bypass tracks and island platforms in a very narrow space if the northbound tracks are on one level and the southbound ones above or below. If one is elevated and the other in a trench, cross roads at grade would be fully grade separated.

    The arrangement would avoid lateral eminent domain takings, possibly even permit existing curves to be straightened out. Unfortunately, many parts of the SF peninsula are a hodgepodge of over- and underpasses and, there are hydrological problems for all tunnels.

  24. I don't see how even with 12tph capacity, you could get away with just two tracks.

    While we might never schedule more than say 8tph, but the extra capacity is great when stupid shit happens. And stupid shit happens all the time. Besides suicides, people clipped at stations, there are unruly passengers that require police escort, the occasional "we overshot the platform" or just a crowd of bikers at various stops is enough to cause the train to be running 10 minutes behind schedule by the time it reaches the end of the line.

    The southbound baby bullets generally need to slow down prior to reaching mountain view about 50% of the time because they're catching up to the local trains by then. I'd say that you definitely want 4 tracks near SF and SJ. If you want a tighter schedule, you'll have to retard the schedule to allow for more recovery time. While that might make the 2 track world possible, you won't get any kudos from either Caltrain or HSR passengers while they wait for each other.

    BTW. Major kudos to caltrain and police after the last suicide:
    They actually left ONE track open despite the incident and had both tracks cleared in 1hr 15 mins. In most cases, they close both tracks, reopen one after an hour and both tracks after 2 hours.

  25. Richard, you're padding the width requirement. Platforms don't have to be 9 meters wide; the New York City Subway uses 5- or 6-meter island platforms when space is constrained. Track centers don't have to be 4.5 meters apart - French LGVs use 4.2, and while French trains are slightly narrower than American trains, LGVs are designed for greater speed and need more clearances.

    This gives 0.5 + 3.8 + 4.2*3 + 3.8 + 0.5 = 21.2 without platforms, and 0.5 + 3.8 + 4.2 + 1.5 + 5 + 1.5 + 4.2 + 3.8 + 0.5 = 25 with an island platform. With 4-meter side platforms, also common in New York, make it 4 + 1.5 + 4.2*3 + 1.5 + 4 = 23.6; you save width by using the side platforms as sound barriers and anchors for catenary.

  26. Full grade separation is a key part of CHSRA's equation in the program EIR/EIS - its what allowed them to proclaim that this Peninsula route had no negative impact (because they conveniently traded off the truly ugly impacts of HSR with the 'benefits' of full grade separation.)

    Now if they come in and suggest NOT doing the grade seprations, opens quite a can of worms on the route decision itself. In fact it would pretty well negate the program level EIR.. THAT would be challenged until Obama's great grandkids were running for office.

  27. Looking at some of the alternatives down on the Orange County segment CAHSR has as one option

    For narrow sections, tunnels..FOR HSR ONLY..Metrolink/Amtrak stay as is. The nimbys crying about tunnel
    tunnel may very well get this option..Caltrain at grade with no grade protection and HSR IN THE tunnels ..far less than digging 4 tracks worth and towns will have no grade improvments. this was aslo noted in the option

  28. I don't see how even with 12tph capacity, you could get away with just two tracks.

    For short stretches they could easily run 20 trains per hour in each direction. Railroads all over the world do it. Even Amtrak and NJ Transit manage more than that in the North River Tunnels.

  29. Clem asked if there are other creeks downtown. Well, not downtown but just south of the Hayward park station there is a creek that runs parallel to tand just along he tracks for a couple of blocks and then southeast through the Fiesta Gardens neighborhood. Water tables are pretty high all around here and there's been a huge effort to deal with flood zones since the the flood maps have been redrawn to include most of the flatland in San Mateo in 100 year food plain. The thrust of the work has been widening and deepening creek channels I think.

    No comment on the noise issue I see...This is a big problem, everybody here sees what happened to Millbrae SSF and San Bruno with the airport and I don't think the Hillborough people are gonna just say "sure we'll settle for new windows" (like those cities did with SFO) ....

    And, just for full disclosure, I think that running trains all the way into SF is just a vanity project for the city. I think HSR's main use will be commuting from the Central Valley to the urban centers north and south and that for that purpose stopping at San Jose or running to the east bay make more sense.

  30. @ Karen -

    railway noise is a complex subject that really deserves its own post on this or another blog.

    In a nutshell, there are two factors to consider: the noise level and the spectral distribution.

    The relevant metric for the level is the so-called sound exposure level (SEL). It combines the measured sound pressure in dB(A), the duration of noise events and their frequency into a single "equivalent" number for the purpose of environmental assessments.

    The spectral distribution of noise is relevant for psychoacoustic metrics such as loudness and harshness. By and large, modern electric trains running on good rails tend to score better on this than motor vehicles.

    Both types of metrics are tricky in that the psychological and physiological effect of noise varies from one person to another. For example, some people can become accustomed to occasional loud background events like trains passing by and stay focused on their task. Others find it difficult to concentrate, let alone sleep, because their brains never learn to ignore the events.

    Noise sources from electric trains include rail-wheel contact, aerodynamic drag, pantograph-overhead catenary contact, transformer, inverters and electric motors. Of these, rail-wheel contact and aerodynamic noise are the dominant sources for fast trains.

    HSR systems minimize both at source by design: the rails are continuously welded ribbons, the contact surface kept smooth by relying on recuperative braking supplemented by disk brakes. Surface roughness is a key factor in contact noise, yet another reason why freight trains need to be kept off fancy HSR rails. The trains also feature optimized nose cones and minimized spaces between cars to keep drag down.

    In addition, noise transmission can be reduced using sound walls and/or the special windows you refer to. Compared to the present Caltrain situation, HSR and Caltrain electrification are expected to bring the following changes:

    - sharply reduced noise level per event due to elimination of grade crossings (no horns, no bells), diesel engines, improved rail bed and modern rolling stock. Freight trains will still rattle, but at least there won't be any nighttime horns and bells.

    - sharply increased number of events by 2030 (up to 10 Caltrains plus a similar number of HSR trains per hour during peak periods, each way). Number of freight trains not expected to change, though UPRR has not committed to that.

    - wider lateral dispersal of noise if tracks are elevated, since most structures next to the railroad are single-story. However, the width of the impacted zone depends critically on many factors, including train design, maintenance of rail and wheel surfaces and the presence of sufficiently tall and effective sound walls. In addition, the zone is narrower where trains run more slowly or there are tall buildings adjacent to the tracks.

    It's fair to ask HNTB, the consulting outfit CHSRA has hired to conduct the project-level EIR/EIS work in the SF peninsula, for some hard data and SEL maps derived from computer simulations. Those need to be compared with maps of the current baseline situation produced using separate simulations, calibrated by measurements in the corridor. The effort involved is not trivial, it would take several months and a fair chunk of change to do properly.

    However, prop 1A(2008) allows CHSRA to spend up to 10% of the $9 billion reserved for HSR on project-level EIR/EIS work and preliminary engineering. Quantifying the noise issue ought to be a high priority. At this point, it's not even clear that the SEL would go up at all. Caltrain and UPRR operations are actually very noisy today, a fact that is often discounted.

  31. Clem, just FYI, but disc brakes actually don't help rail smoothness. With tread brakes, the wheels are constantly being polished into a rounder shape, while with disc brakes (and drum brakes) you don't get that polishing effect and the wheels become bumpier, and make the tracks bumpier in a sort of mutual reinforcement process. I know this is a problem on street railways, no idea to what extent it's a problem for HSR.

  32. @ arcady -

    streetcars have to brake far more frequently and abruptly than HSR trains. Rail-wheel noise is a secondary consideration in urban traffic and, purely recuperative braking is often not possible.

    In general, tread brakes actually increase surface roughness, especially if there is no ABS system to keep wheels from seizing up. Over time, rough wheels also rough up the rails. Note that we're talking about a microscopic phenomenon here, the naked eye would have a hard time telling a rough rail from a smooth one.

  33. Rafael: I might be mistaken, but these are observations I've heard from people working with street railways with PCC-like cars, and I suspect that they're right. Tread brakes smooth out the wheels by simply wearing away the surface. The issue of flat spots produced by locked brakes is actually somewhat orthogonal to this, and tread brakes can't really help with that. But without tread brakes, minor imperfections in the wheels tend to imprint themselves on the rails, which imprint them on the wheels of successive trains and so on, resulting in a sort of wave on the rail head. I think this might even be visible to some extent if it gets particularly bad (look closely at street railway tracks some time), and you can certainly hear it. The way to get rid of it is by regrinding the rails and re-cutting the wheels.

    Come to think of it, didn't BART have this problem? I recall hearing something very similar to this in a press release announcing their new rail grinder which was meant to reduce wheel-rail noise. I'm pretty sure BART trains don't have tread brakes, and they're definitely astoundingly loud.

  34. Oh man. Giant steaming helpings of Male Answer Syndrome.

    If you don't have a clue about rolling contact wear, or about tribology, or about brake system design, or about wheel-rail interaction (a very active field), or about railway noise mitigation (a very active field outside North America) then the correct action is ... to say nothing.

  35. Anonymous @13:12 -

    Do you actually have anything constructive to add, or is your intent just to show superiority without having to do any work?

  36. Timote, he's not even showing superiority. Showing superiority involves giving some useful information. Richard does it, a lot. Anon here is just asserting superiority - for all we know, he could be an ornery high school student who heard a few buzzwords in conversation.

  37. OK, settle down... I think what sent anon ballistic was a mention of 1930s PCC streetcars in the context of a modern, active research & development field. It jarred me too.

  38. Who's to say I didn't post that anyway? Anybody can snipe anonymously. For that matter, anybody could pretend to post as me.

  39. Anon here is just asserting superiority

    I stand corrected.


  40. Train Watcher13 July, 2009 19:47

    @ Clem

    Any chance you can go to a PCC meeting and talk to the group about what is realistic? Your posts regarding Tony C. on the Town Square article regarding his "tunnel vision" were excellent. We need your wisdom and experience. Can you come?!

  41. @Karen Re: noise, did you miss where Adirondacker posted the following?

    Electric trains are much quieter than diesels, the grade crossings along with their noise, will be eliminated and thw whole Peninsula will be getting faster train service.

  42. The three stations in San Mateo generate more ridership than San Jose. OOOOkay, so what's your point Clem? Or is this just your way of getting another cheap shot in on the Bay Area's largest city and economic engine?!

    Anyhow, that statement won't be true in 10 years.

  43. Oddly enough, I recently spent an hour -- RUSH HOUR -- 5:30 to 6:30pm -- at San José Diridon Station last week; the hour when tens of thousands of workers leave the humming office towers of San José, Capital of Silicon Valley, and head home via efficient VTA, Amtrak, ACE, and Caltrain services.

    My observation was that typical boardings on Caltrain were between 10 and 20 passengers. I believe the usual industry term for this is "non revenue service".

    The multi-modal VTA light rail connection was even more stunning: average boarding per train of just over 1 person during the time I observed the platforms (which was basically any time a Caltrain wasn't departing.)

    As for the transit oriented development condos around the station, let's just say I didn't see a single pedestrian.

    So by all means let's keep building on this success! It's clear the station isn't anywhere big enough or multi-modal enough to deal with the demand of the tenth largest city in the most important country on the only known habitable planet in the universe. The arrival of BART and HSR will make San José Diridon Station so multi-modal I get dizzy just thinking about it. Light rail boardings might rocket up to dozens per hour, while the 4200 BART seats per direction per hour (10 car trains every 6 minutes) may have have occupancy as high as 1%.

    Meanwhile, San Mateo's downtown isn't a bad place to visit at all. It isn't up there in the major metropolis league with Palo Alto or Burlingame, but there's stuff to do and stuff to buy. Check it out!

    None of this is just prejudice or "hating on" San José, Capital of Silicon Valley: it's more looking at the numbers, taking the time to see what's happening in the world, rather than on a blog or in the propaganda pieces of local politicians.

    PS I work in Silicon Valley, have worked in Silicon Valley for more than a decade, I can make the payments on my mortgage thanks to the largesse of the Captains of Industry of Silicon Valley, but I have no reason to ever visit San José, Capital of Silicon Valley, nor do most of my fellow Silicon Valley serfs.

  44. According to Caltrain San Jose sees 2,983 boardings per day, making it the fourth busiest Caltrain station.

    So you should have seen about 30 riders board each train on average.

    "the hour when tens of thousands of workers leave the humming office towers of San José, Capital of Silicon Valley"

    When I was in San Jose during a non-holiday weekday I didn't even see this. Downtown felt like it was deserted. Few cars and even fewer people. Even Downtown LA has people in it during the day.

  45. I rode VTA Light Rail during rush hour on a non-holiday weekday and tried to figure out what was wrong with it.

    One, it's trying to be a quaint little streetcar for downtown while simultaneously trying to bring in commuters from the suburbs. It's slow and I have to imagine it pisses off people trying to get to work who are coming from further away. Maybe I'm wrong, but I would remove half of those stops in the middle of the Alum Rock-Santa Teresa line.

    The Green Line toward Mountain View felt like it had too many goddamn stops too, and a meandering route. These stations really need to be at least a mile apart.

    Two, there were portions of the blue line toward Santa Teresa that run in the median of the freeway. Freeway stations suck ass. They are often difficult to get to and unbearable to wait at. I would bet this contributes to the low ridership problem. I think the Green Line in Los Angeles shares a similar issue. But then again there are a hell of a lot more people on it than on VTA light rail.

  46. Anon, San Jose is a bedroom community to the cities to its north. You should check Caltrain boardings at 8 am, not 5:30 pm.

  47. Basically, the way the population and jobs in Silicon Valley work out is that the population is in the south, and the jobs are in the north, with minor job centers scattered elsewhere (Cupertino, Downtown San Jose). Palo Alto, Mountain View, Sunnyvale, Santa Clara, each probably have around the same amount of high-tech office space as North San Jose (especially Santa Clara). The reason the VTA light rail fails is that the somewhat circuitous and very, very slow route through Downtown makes it much less useful for commutes from south of downtown (where the people are) to north of it (where the jobs are). Add to that the fact that the many of the jobs are in fact even further west in Santa Clara and Sunnyvale, and many are not on the light rail, and it makes for a very long trip. Most of Caltrain's Silicon Valley-bound commuter ridership is going to Palo Alto or Mountain View, most of ACE's ridership takes the shuttle buses at Great America (the platform there is PACKED at rush hour). And even light rail gets decent loads during rush hour on the lines south of Downtown (which are at least fast).