23 May 2009

The Shape of Palo Alto

Several years ago, the city of Palo Alto commissioned a public art work entitled The Color of Palo Alto. At considerable expense ($75,000), artist Samuel Yates set out on a techno-artistic quest to determine the exact shade of Palo Alto. In a similar vein, and bearing in mind that artistic form is more than just color, we set out to determine the shape of Palo Alto.

Why does the shape matter? While the area around the tracks in Palo Alto is pretty flat, small variations in the terrain do have an effect on how the high speed rail project will be configured as it crosses through town. The vertical alignment of the tracks must not only conform to the surrounding grade, but it must be routed over or under every creek, road, pedestrian or bike crossing while conforming to acceleration limits imposed by passenger comfort as well as train stability and performance.

There have already been strong hints as to what shape this vertical alignment might take, as previously discussed in Focus on Palo Alto. As part of its program-level EIR/EIS, the California High Speed Rail Authority produced a set of maps of the Caltrain corridor that fueled a firestorm of controversy in Palo Alto, with residents worried that raised tracks would form a Berlin Wall and divide the city. While the CHSRA now professes to consider all vertical alignment options for its project-level EIR/EIS, ranging from tunnels to elevated tracks, there aren't that many feasible and affordable ways to design the vertical alignment of HSR tracks through Palo Alto-- and it turns out that the shape of Palo Alto can tell us a lot about what those are.

Working from Caltrain track survey data, here is the shape of Palo Alto, with the vertical scale greatly exaggerated: (click to expand)


Note a few salient features:
  • Palo Alto isn't as flat as you might think. Who knew that the University Ave station is nearly 40 feet higher than the California Ave station?

  • There are only four grade crossings in all of Palo Alto

  • Including overpasses, underpasses and grade crossings, there are only nine places where pedestrians and bikes may cross the tracks in all of Palo Alto, and only seven places where roads cross the tracks

  • The entire southern half of Palo Alto has only two places where pedestrians can cross the tracks, and both of them are very busy intersections.

  • Four creeks cross under the tracks
As we consider various grade separation options, keep this in mind: while grade separations are a sine qua non condition for high speed rail, they could also become necessary in other non-HSR scenarios for the peninsula. For example, if Caltrain traffic increased to the point that grade crossing gates blocked rush hour traffic for too long (such as if HSR were to terminate in San Jose), or if BART were to connect Millbrae to Santa Clara, as originally planned in the 1960s. Grade separated tracks are by no means a unique requirement of HSR; they are likely to be required in Palo Alto within the next few decades, HSR or not.

Constraints on Vertical Alignment

Before evaluating various vertical track profile options in Palo Alto--especially if you're going to have a charrette--it's important to lay out a few simple, fundamental design rules that govern what you can and can't make the tracks do.
  • The maximum grade (how steep the tracks can be) will be limited. High speed trains and electric commuter trains can easily climb 2% or 2.5% grades. Freight trains can't climb as steeply, and work best at 1% or maybe 1.5%.

  • The radius of vertical curves (humps and sags) is constrained by maximum train speed. The vertical curves have to be gentle, unlike a roller coaster, to keep passengers comfortable and keep the train stable on the tracks as it crests over a hill. For 125 mph (200 km/h) operation, here are some typical track design standards:
























    Standard (at 125 mph)
    Typical RadiusMinimum Radius
    European Technical Standard of Interoperability14 km (9 mi)10 km (6 mi)
    Germany16 km (10 mi)10 km (6 mi)
    Sweden17 km (11 mi)10 km (6 mi)


    Some standards do allow vertical curves with a radius barely over 6 km (4 mi), but only in exceptional circumstances. For the peninsula, using the minimum value of 10 km is probably a good starting point. This radius results in a vertical acceleration of 0.3 m/s2, or about 1/3 to 1/4 of the acceleration you might feel in a typical elevator--infinitely smoother than a roller coaster!

  • To the extent possible, the vertical profile should avoid obstacles like creeks or existing overpasses and underpasses. While earthworks and concrete can be used to solve any such puzzle, the cost may be prohibitive.

  • Where roads cross under the tracks, the vertical distance between the road surface and the top of the rails must be 20 feet to clear trucks and buses.

  • Where tracks cross under roads, the vertical distance between the road surface and the top of the rails must be 30 feet to clear freight trains and overhead electrification.
The chart at right shows vertical track profiles for various combinations of these parameters. On the top half of the chart is a 20-foot elevated grade separation, with representative 1% or 2% grades and 10 km or 16 km vertical curve radius. On the bottom half is a 30-foot depressed grade separation (trench), with the same variations of grade and vertical curve radius. The chart gives you an idea of how long these structures need to be, based on how high (or deep) they are.

The Elevated Scenario

Elevating the tracks over Palo Alto's few remaining grade crossings is one scenario being proposed by the California High Speed Rail Authority, and is likely the cheapest.

Taking a step back: the only reason to elevate track that isn't already elevated is to lift the track over a road or pedestrian underpass, also known as a grade separation. The only place where that might happen is in the vicinity of grade crossings. Elevating the track anywhere else would serve no purpose and is not being contemplated.

Here's what the elevated scenario might look like:


  • All vertical curve radii are 10 km (6 miles)

  • At Charleston and Meadow, the tracks are raised by 7 feet and the road sunk by 13 feet. This embankment height was given in the CHSRA's program EIR/EIS maps.

  • At Churchill, the tracks are raised by 15 feet, with 1.5% ramps. The CHSRA's program EIR/EIS maps showed the embankment starting well north of Palo Alto high school, presumably to avoid dipping down and back up; however, there is absolutely no harm in the hilly profile shown here. The only people who might dislike it are freight train operators, not HSR, because of slack action. Speaking of freight trains, this embankment could be made even shorter and steeper if it didn't have to be designed with the gentle grades required for freight trains.

  • Alma (a.k.a. Palo Alto Ave) is lowered to cross under the existing track level.

  • As evident in the diagram, any claims of a continuous 20-foot wall bisecting Palo Alto from one end to the other are exaggerated.
The Depressed Scenario

With respect to the outcry over the community impact of raised embankments, it's worth looking at another scenario... and it doesn't involve tunnels! While tunnels are often contemplated as the only possible alternative to embankments, it is possible to grade separate the Palo Alto crossings by depressing the tracks into trenches. Trenches are preferable to tunnels because they can be narrower and involve far less earth moving and concrete works... although still far more work than embankments.

Here is what the depressed scenario might look like:



  • Note the length of the grade separations increases considerably, since they must dive a full 30 feet under street level and do so with gentle, train-friendly grades. The excavation itself would be even deeper, since the red curve is drawn at the top-of-rail level. Do note that accommodation of freight trains requires an extra 3 feet of vertical clearance.

  • The trenches are shown with 1.5% ramps. Shallower ramps are not possible, since the natural grade is already nearly 1% near the Churchill crossing. Steeper ramps (2% or 2.5%) would be possible if it weren't for those freight trains. (Do we see a pattern yet with those pesky freight trains?)

  • At Churchill, the northern ramp would not clear the existing Embarcadero grade separation. This complication might be solved by locally depressing Alma to match the track level, and rebuilding Embarcadero as a low overpass rather than an underpass--but as always, the devil is in the details.

  • At Charleston and Meadow, the trench would encompass the existing location of the Barron and Adobe creeks. Grade-separating the creeks is possible, but it involves lots of concrete, pumps, and ongoing maintenance expenses.
While the depressed trench scenario introduces several challenges, they pale in comparison (and cost) compared to building a full-fledged tunnel.

The Big Picture

Despite the recent controversy in Palo Alto, the city is not even close to being the most difficult to build high speed rail. There are far more challenging puzzles up and down the peninsula, for example downtown San Mateo. Many of the same design principles apply to any location on the peninsula, and as time allows we'll start discussing some more of these puzzles. As for Palo Alto, the squeaky wheel gets the blog grease!

46 comments:

  1. Nice work as usual.

    Re "The only place where that might happen is in the vicinity of grade crossings. Elevating the track anywhere else would serve no purpose and is not being contemplated.":

    I claim that's not strictly the case.

    At grade tracks, such as the existing Caltrain line, form a considerable barrier to pedestrian movement. Existing grade crossings -- their expense, their small number, and their physical design -- are largely determined by the perceived needs of motorists, while existing grade separations -- especially nasties like Oregon Expressway -- are solely driven by 1950s highway design thinking.

    A modest amount of track elevation can make it possible to create new pedestrian (and bicycle) cut-throughs. The dreaded "Berlin Wall" can in fact re-unite neighbourhoods. Not that motorists care.

    Secondly, a modest amount of track elevation at stations can make for far superior connectivity and pedestrian circulation -- instead of making people descend into tunnels and then climb back up to the tracks, one should aim to allow people to walk through the station, and placing buses and taxis and downtown right at their feet.

    Raising the tracks say 10 feet in downtown Palo Alto could do a lot for downtown. This isn't enough to bring University back up to grade, but it could mitigate some of the cars-before-people unpleasantness of the existing University/Alma mini-interchange. It would also make an Alma grade separation adjacent to San Francisquito Creek much less disruptive.

    In fact, I don't see how to make anything work without doing so. (The completely inconvenient and crazily expensive and inutile VTA bus station on the west side of and completely disconnected from the Palo Alto train station is exhibit A of "not working". You can't get there from here!)

    "Unnecessary" modest track elevation between downtown PA and Churchill could offer to opportunity to mitigate the Embarcadero underpass. Or not. People in Santa Clara County do love their urban freeways.

    ReplyDelete
  2. Adirondacker24 May, 2009 02:40

    Elevating the tracks over Palo Alto's few remaining grade crossings is one scenario being proposed by the California High Speed Rail Authority, and is likely the cheapest.

    Likely but not that much cheaper. You still have to dig deep to put up anything much higher than a curb. Seismic considerations may make the viaduct more expensive - holding up a wall can be cheaper if you can put tie backs into the soil...

    Richard, a minor quibble - the neighborhoods on either side of the track happened after the railroad. They were never united, they therefore can't be reunited. "Reuniting" things plays into the meme of Berlin Wall and division.

    I've never looked very hard at the Palo Alto station. I have thought to myself "Robert Moses would be proud" It's not a mini interchange, it's a full blown cloverleaf, on a small scale that has low speeds but it's got all the parts. And it's right next to a full blown diamond interchange at El Camino Real. All they need is a toll booth and an oil refinery to two and it's the New Jersey Turnpike. . . Nah the designers at the Turnpike could do much better than that.

    The only way I can see to get from one platform to another is across the tracks or taking a four block walk along University. There may be staircases hiding in there someplace but I don't see anything that looks like it might be stairwells. All they had to do was make the underpass just low enough to exclude buses and it could have been designed by Mr. Moses himself.

    Is the existing overpass strong enough for four tracks or will they be ripping it out anyway?

    If they lower the tracks and raise University it makes it much more pedestrian and bus friendly. There can be big open plazas around the station Sell it as a traffic calming method for University and Alma. . . nah they aren't going to give the grade separations between University and Alma and El Camino..

    While wandering satelite and street views I noticed this . I can see where catenary. awful catenary with a few wires here there is going to just ruing the bucolic charms of Palo Alto. . . how tall is that pole?

    ReplyDelete
  3. Richard: "The dreaded "Berlin Wall" can in fact re-unite neighbourhoods."

    Adirondacker: "Richard, a minor quibble - the neighborhoods on either side of the track happened after the railroad. They were never united, they therefore can't be reunited. "Reuniting" things plays into the meme of Berlin Wall and division."

    How so? If the current situation is the Berlin Wall and the elevation punctures holes in it ... is that reinforcing the meme or subverting it?

    ReplyDelete
  4. Clem: do you have a smooth transition from a straight line to a curve of minimum radius? If you do, what's the maximum allowable jerk?

    ReplyDelete
  5. One possibility that should be considered for Churchill is having a car-only crossing (no trucks), with less vertical clearance and thus a smaller elevated structure. I agree with Richard that they should raise the tracks at Palo Alto somewhat, and reconfigure the whole University Ave/Alma/El Camino mess to be less hostile to pedestrians. It would help with pedestrian circulation once the station has four tracks and two island platforms. They might also want to consider raising the tracks at California Avenue to alleviate the flooding problem in the Oregon Expressway underpass and allow for a better pedestrian connection across the tracks. Other than in that area, though, there's not much opportunity for connections across the tracks, because there's pretty much a solid line of houses on the southwest side. The neighborhood were never connected to begin with, so there's not much opportunity for easy connection.

    ReplyDelete
  6. Adirondacker24 May, 2009 12:32

    Bruce, it's a minor quibble. Richard was pointing out that grade separating Caltrain improves things all along the line. But calling it "reunitiing" implies that the railroad doesn't already divide the everyone. Subtly saying that the viaduct will be dividing people. .. who are already divided. Except for the Darwin Award competitors who cross the tracks anywhere, the few places where there are controlled crossings make it difficult to cross the railroad. There's gotta be a better way to express that, better than reuniting or even uniting. One that escapes me.

    Compare "Uniting neighborhoods" with "Reuniting neighborhoods"
    Uniting is neutral or a bit positive. Reuniting implies that there was a heart rending division that has to be repaired.... and the way to sooth this awful division is to get rid of the railroad by tunneling it or moving to 101 or the East Bay...

    I looked a bit harder at the Palo Alto station. If they decide to make Palo Alto the Mid Peninsula station how much of what's there is going to survive? I looked at the bus route maps. I have a feeling that some bus passengers spend five minutes negotiating the cloverleaf, to get to the place where they got off the train.

    They have a chance to do this much much better. Raising the tracks puts the station and University under an overpass. Lowering the tracks and raising University puts pedestrians, buses, cars that pass through out in the daylight. The bus and car exhaust is out of the station. Opens up the station itself to daylight.

    ReplyDelete
  7. "Other than in that area, though, there's not much opportunity for connections across the tracks, because there's pretty much a solid line of houses on the southwest side."

    AddisonKelloggSeale (nice city park)
    California!!!!!!!!!!!
    Oregon HellwayEl Dorado/sad concrete Matadero Creek-SewerEl Verano (requires easement through parking lot; idealistic/unrealistic, I know)
    Ferne/Del MedioIt's harder to think of places not to add cut-throughs, though I'm sure the good automobilist burghers of Palo Alto (we'll be murdered in our beds by brown people on bicycles!) will come up with them.

    Re "having a car-only crossing (no trucks), with less vertical clearance":

    That's simply not going to happen. I'd be surprised if it were legal in the first place; in the second none of the city, county or state transportation agencies would go for it; and lastly there's no way any rail agency would want to erect a structure which is guaranteed to be attacked regularly with massive directed energy weapons. See San Mateo.

    Re "They were never united, they therefore can't be reunited":

    Besides the distant aerial-photo tourism on which this seems based, this sort of comment suggests very dim memory of being a kid without a car seeking to get from A to B. Just because some houses were built as part of two different subdivisions and not one doesn't mean that half mile or worse detours around them are warranted. Draw up the bridge! Prepare the boiling oil!

    Re "smooth transition from a straight line to a curve of minimum radius":

    Think about it for a couple seconds. You're allowed to pencil and paper, but knowing g~=10ms^-2 ought to be enough.

    PS the first three and top 8 out of 10 Google hits for "g 9.8" are not what one might expect.

    ReplyDelete
  8. @Adirondacker: there are three pedestrian underpasses connecting the platforms at University Ave. Two branch off of the sidewalk tunnels of the underpass, and emerge north and south on each platform. The third underpass is at the north end of the station building.

    @Alon: no, vertical transition curves are not included. For comfy jerk values like 1.5 m/s3, the transitions are a couple dozen feet long and result in height differences of a few inches. In other words, for such enormous curve radii, the transitions do not materially affect the dimensions of the grade separation; that's why I ignored them.

    @arcady: I doubt a car-only crossing (i.e. no school buses) would work next to a school... then again, many PA parents do pick up & drop off in a Chevy Tahoe. And you'd have the issue of emergency vehicle access.

    In general, the University Ave / Alma intersection is more a horizontal alignment problem than a vertical alignment problem. When they built the University Ave grade sep in the 1940s, the tracks were shifted west by quite some distance. For details, see the Palo Alto post. And yes, SP built it with four sets of steel girders!

    ReplyDelete
  9. Wait, is this for real? Palo Alto is freaking out over a total of 1.5 miles of raised tracks - 2 max? Seriously, learn to pick your battles! If the City of Palo Alto wants to spend tons of tax dollars studying and complaining about the impact of HSR, why not redirect it to finding an architecturally artistic design for the grade separations? Or how about lobbying for a rebuilt/improved transit center and train station?

    ReplyDelete
  10. Richard: I know that g = 9.8 m/s^2, thank you very much. I'm asking about jerk. In one of the previous threads, you talked about line-spiral-circle-spiral-line transitions; what I'm interested in is the jerk in the spiral.

    ReplyDelete
  11. Adirondacker: "Bruce, it's a minor quibble. Richard was pointing out that grade separating Caltrain improves things all along the line. But calling it "reunitiing" implies that the railroad doesn't already divide the everyone."

    It might not be so minor, given that the main competitive advantage of the HSR opponents is the seeming inability of HSR advocates to engage in effective advocacy.

    Calling it "re-uniting" quite clearly implies that the railway line presently divides people and whatever option is "re-uniting" is therefore prima facie superior.

    Standing against that, its inaccurate, opening it to attack by quibbling. And as well, calling it "re-uniting" also clearly implies that the present railway line is "bad railway line, dividing those previously united neighborhoods".

    And there is also the question of how much to focus in on the area in question ... its only between University Ave. and Embarcadero that the option is available.

    Rather than arguing the priorities in theory, it should be argued by presenting plans that incorporate the different priorities in practice.

    @ arcady ... yes, that's an important point. Its not much more than a mile between Embarcadero and the Oregon Expwy, there should be some consideration of whether a full semi-trailer height overpass is actually required. If the height of the overpass is dropped by 5ft., then gradient can be dropped to 1:100 and still have a smaller footprint.

    ReplyDelete
  12. What's the overhead clearance required for a school bus or firetruck? Would 12'6" be enough?

    ReplyDelete
  13. "I'm asking about jerk [entering a vertical curve]. In one of the previous threads, you talked about line-spiral-circle-spiral-line transitions; what I'm interested in is the jerk in the spiral."

    How much spiral would you need for a horizontal curve of radius 10 km?

    ReplyDelete
  14. Adirondacker25 May, 2009 00:50

    Besides the distant aerial-photo tourism on which this seems based.

    I've been on the Peninsula many times. The Peninsula is frighteningly similar to the places I've lived on the East Coast.

    I lived most of my life in Newark NJ and it's suburbs. I compare my experiences in New Jersey to what I experience in California. The similarities are striking.

    What happened in railroad suburbs on the East Coast and Midwest isn't all that much different from what happened on the Peninsula. There's railroad suburbs all over the world that work just like the Peninsula. The architecture and landscaping may be a bit different but they are all very very much alike.

    For years I was all over the metro area for work. I have lots of experience comparing what's on the street map to what's outside the windshield of my car. When I've done the same thing in California I've had the same kind of experience.

    this sort of comment suggests very dim memory of being a kid without a car seeking to get from A to B.

    Once I was old enough to cross streets by myself if I wanted to go someplace I walked or when I learned how to bicycle, bicycled . If it was too far to walk or bicycle there was a rich network of buses. I'd go months without being inside a car. 50 cents, and the world was my oyster. A quarter for bus fare to get there and a quarter to get back. I didn't own a car until I was 20, didn't need one. Even then I didn't need one, it was a luxury I could afford.

    ... and I've never wanted to get from A to B. I always want to get from here to there. :-)
    ... and very frequently with railroads and highways crisscrossing all over the place in metro NYC you can't get there from here.

    Just because some houses were built as part of two different subdivisions and not one doesn't mean that half mile or worse detours around them are warranted.

    I'm not suggesting that. Just that calling all the good that happens, "reuniting". reminds people that they aren't united. Which leads some of them to "bad railroad" as Bruce wrote and coming to the conclusion that the only solution is a tunnel so that the vicious railroad that has been preventing the arrival of the Post Millennial harmony and peace, is gone.

    I'm sure there are official documents floating around that have phrases like "improving community interconnectedness by facilitating pedestrian interchanges across the corridor" To rational people who speak English and not DOTese, reuniting is a succinct way to say that. Unfortunately it also has bad connotations. Berlin, which comes up frequently in these discussions, was divided and then reunited. ... there's gotta be a better word.

    I'd be surprised if it were legal in the first place.

    I ran that past my NJ Transit source and my NYDOT sources. They say it's legal.

    in the second none of the city, county or state transportation agencies would go for it.

    Both sources then said that Federal government isn't going to fund anything lower than 14'6" using conventional transportation funding. So while there isn't an explicit prohibition, you aren't going to the money to build it, it wouldn't conform to standards. Both then theorized that there may be some sort of highly unusual circumstances when something less than 14'6" might get approved. Offending the delicate sensibilities of Palo Altans wouldn't be one of those extraordinarily rare circumstances. Especially if there are alternatives that do comply with a 14'6" clearance.

    Fun to conceptualize but it's not going to get serious consideration.

    And yes, SP built it with four sets of steel girders!

    Which made eminent sense in 1940. Makes sense when Caltrain does it today.

    I reread the Palo Alto post. Do those four girders align with high speed tracks? Have they been maintained for the past 70 years? Which makes me wonder if they are willing to use 70 year old girders even if they have been well maintained. Is it all up to seismic code? .... How much of what's there now is going to survive?

    ReplyDelete
  15. @ arcady ... good question, clearly 9 feet is not enough.

    I gather from a San Joaquin County document (pdf) that the California fire equipment standard is 13'6".

    The Interstate standard is 16' in rural areas, 14' in urban areas, and at least one urban routing of 16'. As a national standard, that has a big impact on designs of heavy motor vehicles.

    What is the height from the bottom of the overpass structure to the top of the rail? Is that in the EIR/EIS documents somewhere? I'd presume that the 20' vertical distance from the road surface to the top of the rail is for a 16' road clearance. Dropping it by 2' would primarily affect semi-trucks, dropping it further would start to affect larger local urban vehicles.

    ReplyDelete
  16. What is the height from the bottom of the overpass structure to the top of the rail? ...

    Typically 5 to 6 feet for a reinforced concrete structure.

    ReplyDelete
  17. Correction: 4 feet. See Caltrain's vertical profile of San Bruno, page 4 (warning: 30MB file), with some detailed grade separation examples.

    ReplyDelete
  18. If it was too far to walk or bicycle there was a rich network of buses. I'd go months without being inside a car.

    Andirondacker -huh? I don't get what you're going on about, but what you're describing here has nothing to do with the Peninsula. A freshman in South Palo Alto can't even take a bus to high school. Not sure what your point was. That you don't know anything about this area?

    ReplyDelete
  19. Clem said...
    "Correction: 4 feet. See Caltrain's vertical profile of San Bruno, page 4 (warning: 30MB file), with some detailed grade separation examples."

    Thanks for that, that elaborates the design envelope for existing level crossings. I was thinking the 20' road top to rail top meant 16' clearance and 4' structure.

    Presuming reinforced concrete is mandatory, since it can be dressed up in ways that steel spans cannot be, there's only 1'6" leeway, for which the game would not seem to be worth the candle.

    (Note that the 14'6" minimum is the same Federal Interstate Highway standard ... the 6" is for laying of fresh asphalt without grinding down the existing asphalt to the subsurface.)

    Of course, a viaduct that happened to be turned into an overpass by making a new low-overpass road connection after the fact would not be under the same constraint, nor would a cycle/pedestrian overpass, but that is only a real consideration for NW of Embarcadero.

    In the vicinity of Churchill Ave., easing the slope to 1% on the SE side would simplify putting in a Peers Park pedestrian and cycle underpass crossing ... the only place other than existing crossings where the roads line up SE of Embarcadero is Cal Avenue, which already has a pedestrian underpass.

    Given the terrain, the viaduct option seems like it would start halfway between El Palo Alto and the Alma St. (Palo Alto Ave) grade separation, rise to an 80ft. elevation, take that through to University Ave. Station, when it is 10ft above grade, then remain 10ft above grade until a shallow rise to cross Churchill. And past Embarcadero, it might convert to a filled wall along Alma, so that a landscaped embankment can face PA HS and Peers Park.

    ReplyDelete
  20. @Adirondacker Oh, it gets even better. Those unfamiliar with Shallow Alto may not realize that Alma St - the street that runs adjacent to Caltrain - has ~50 ft tall utility poles on it for almost the entire length of the city. These poles are far more visually intrusive than catenary poles (which need only be ~30 ft...cf New Haven-Boston electrification project), but for some reason the city has yet to spend the minimal sums required to bury these horrible aesthetic scars that permanently blight this otherwise pristine neighborhood (that last part is meant somewhat sarcastically).

    Question: Why should California taxpayers spend up to $1,000 million to protect Shallow Alto from the blight of 30' overhead wires and two short stretches of retaining wall (max height 7' for one section and 15' for the other section) when the city isn't even willing to spend just $10 million to eliminate the blight of 50' overhead wires?

    Answer: Because when the state pays, it's free, but when the city pays, you have to actually consider whether the benefits are worth the costs. Apparently the City of Palo Alto does not consider the benefits of undergrounding 50' overhead wires to be worth even a scant $10 million.

    ReplyDelete
  21. But Mike, you are forgetting that one of those retaining walls and/or embankments (given the width, if there is a retaining wall on the Alma Street side, there would seem to be space for an embankment then an embankment down to a lower retaining wall) behind the bleachers at one side of a sport field.

    And there is no way that the students of Palo Alto's High School could possibly concentrate in their classes knowing that there is are ominous, scary trains looming over a sports field.

    ReplyDelete
  22. Bruce, Yes, good point. Fortunately those are the opposing team bleachers, however, so in that context the scary trains should be viewed as a benefit rather than a cost!

    ReplyDelete
  23. Once again Clem has presented a lot of important information in a very useful capsule. It does put the whole "OMG Berlin Wall be afraid!" outcry in a completely different light. Two short-ish elevated sections for grade separation at Charleston, Meadow and Churchill, and a road underpass at Palo Alto/Alma are a far different result than the fearmongering about a 20 foot high, 75-foot wide concrete slab running the entire length of the ROW. It is basically the same result as the grade separations that are inevitable with Caltrain 2025, and much harder to argue against.

    I think, in the totality of the circumstances, selling an elevated section as "uniting" or "re-uniting" neighborhoods is going to get drowned about by the faux outrage over "Berlin Wall!" Just getting the grade separations will do a lot to enhance ped/cyclist mobility and safety. I'm not sure that the enhanced mobility on offer by modest elevation is going to be a strong selling point to people worried about HSR. Changing things as little as possible may well be a much easier sell.

    And can I just say that if I were a student at Paly I'd be mortified that so many people think Paly students are such precious flowers that they can't tolerate an elevated railway next to their athletic fields. Please.

    ReplyDelete
  24. Nice work!
    For the Elevated scenario, at Charleston and Meadow, I would think that because Alma is so close to the ROW Alma would have to be raised as well as the tracks for the 13' depth to work. It might be better to go the full underpass route in which case no track or Alma street elevation would be required. The interconnections between Alma and Charleston/Meadow could be done somewhat similar to the interconnections between Shoreline and Alma in Mountain View.
    For the Depressed scenario, filling in the Embarcadero road underpass would have some advantages. Embarcadero could be straightened, and widened to five lanes (2 east, 2 west and one left turn). Alma could be widened in a similar fashion. A really smart traffic signal could manage the (at surface level) intersection. Maybe even some grass and shade trees. But in order to do something like this, the tracks would have to stay deep longer so the trains would be fully underground as they crossed Embarcadero.

    ReplyDelete
  25. Note that California Avenue station sits in a saddle.

    With just a little bit of track raising (ie levelling, so it doesn't go quite as far down and then up again) it would be possible to (a) have direct connection to and though the (central, island, naturally) station platform from both sides of the station and (b) re-connect severed California Avenue east to west. Roughly like this.Not that that accords with fantasies of grade separated road junctions, mini flyovers and coverleafs, Alma as a wanna-be freeway, and all the usual Santa Clara County cars-before-everything insanity, of course. But in most civilized parts of the world people would leap at such an opportunity.

    And in downtown Palo Alto, Lytton or Everett would continue, with only minor grade change, though the station (bus and taxi magic intermodalism), to the west. Roughly like this. Very nice, indeed! Such an opportunity to do the right thing! So certain never to happen!

    ReplyDelete
  26. Joseph said...
    "For the Elevated scenario, at Charleston and Meadow, I would think that because Alma is so close to the ROW Alma would have to be raised as well as the tracks for the 13' depth to work."

    Why raise it? For the split grades where the drop is shorter than the rise, why wouldn't Alma drop to a regular intersection rather than the expensive and land-hungry mini-freeway sculpture.

    ReplyDelete
  27. Richard:
    Ask people in those Palo Alto neighborhoods which is more important: "reconnecting" Lytton and California, or build the elevated railway structure as depicted in the simulations, and I think you know what answer you will get. (Hint: I don't know anybody who has a burning need to walk to El Camino Real.)

    One thing that might help is to not make the elevated structures in the simulations look so much like, well, the typical Santa Clara county freeway. Here again I think the freight requirements are a real detriment. The higher load capacity requires bigger and more columns.

    ReplyDelete
  28. Re Ask people in those Palo Alto neighborhoods which is more important [...]:

    Yes, well we all know the answer to that, everywhere. (Build nothing.)

    But the fact is that a four track, four platform face station in Palo Alto (or a four track, two platform track station at California Avenue) is going to fill out the right of way almost from edge to edge. (Here and elsewhere.)

    Add to that the 400m long platforms hat are appropriate to downtown Palo Alto, noting that even regional Caltrain traffic will eventually justify 300m trains, HSR or no. The station footprint is inherently large.

    And the station isn't going underground, guaranteed.

    A good question then is how people are supposed to get to and from the trains, or get around the train station even if they have no business there.

    You'd think people might care about that sort of thing, just a teeny bit. (I do know that anybody who attempts to get to or from the northbound platform at the brand spanking new Caltrain-engineering-designed California Avenue Station would like to crucify whoever was responsible for making it as hard and slow as possible to get to the trains.)

    The alternative to a modest (modest!) amount of elevation is to simply build impenetrable blank security walls hard up against existing streets and sidewalks, extending for a large fraction of a mile. You can hear the trains, but you can't get to them!

    As for the quality of the simulations, I understand there are people paid millions a year to do this shit. I'm not one of them. Feel free to help out. My Sketchup models can be yours.

    As for columns and spacing, what's shown is 15m (50-ish foot) spacing, and even at BART weight you're not going to get much less than that around here. I try to be honest, which is why I'm a failure.


    PS: Here's a better photosimulation. I worked hard at this, even plopping in a scary freight train -- I do hope everybody appreciates the effort. Like, where's the Berlin Wall at?

    Or how about this? (Many more of that stunning project.)

    Happy?

    ReplyDelete
  29. For comparison, here's some 1950's-era american railroad design. That's the Long Island Railroad at Rockville Centre, a line used by both freight trains and 12-car commuter trains. The station has two tracks and an island platform. For Palo Alto, imagine two such things next to each other, for California Avenue imagine this but with two more tracks on the outside. (and yes, I have come around to seeing the benefits of FSSF: the ROW is wide enough for it, it works better with short sections of triple track and speeds of 125 mph, and makes freight operations a bit easier, since in the evening, the freights can use the outside tracks to access sidings while the passenger service runs on just the inner two tracks. Also, Richard's ideas for Palo Alto and California Avenue are good ones: basically, it makes the station the connection between the neighborhoods on either side of the tracks.)

    ReplyDelete
  30. Arcady: freight trains are too heavy to use the HSR tracks, which are intended to be used only by trains under 17 tons per axle. Heavier trains will wear out the tracks too much, increasing maintenance cost (of which Union Pacific won't pay a dime).

    ReplyDelete
  31. Alon, I highly doubt that there will be special HSR-only tracks. The official plan is four tracks fully interchangeable between Caltrain and HSR. And 125mph passenger traffic is certainly not incompatible with freight trains, at least not in Rhode Island. On actual high speed lines with speeds above 125-150 mph, yeah, freight trains will have to be completely excluded, but the Peninsula Corridor is not really a high speed line.

    ReplyDelete
  32. Arcady, Caltrain is going to run lightweight EMUs; it won't have the same effect on the tracks as freight trains. At any case, heavy freight won't make the tracks unusable by HSR - it will just increase the cost of maintenance.

    ReplyDelete
  33. Remember, we're talking about approximately three (in the bright future of freight maybe five) freights a day, of which about two will be "heavy": the 70 or so carloads of gravel that are already running on the corridor, and the train of doublestack containers or car carriers that thus far exists only in the imagination of the Port of SF. Also, the speed limit for freight is currently 55 (and they want to reduce it to 50 for signal reasons). Compared to 150 Caltrain runs and 50 HSR runsat up to 125 mph, which one is going to put more wear on the tracks? Actually, I don't really know, I'm not an expert on track structure or dynamics, and I'm betting neither are you, but I think it's not so obvious that the big bad freight trains will be destroying the track.

    ReplyDelete
  34. I'm not an expert, but I do know that track wear is proportional to the fourth power of axle load (link). Apparently, reducing speed does not lessen this effect: the South Dakota DOT in fact claims that "Road damage increases significantly when heavy vehicles are driven more slowly." A 70-car freight train with an axle load of 25 tons will cause as much track damage as 40 half-length HSR trains with an axle load of 17 tons.

    ReplyDelete
  35. I somehow think that pavement dynamics are very different from wheel-rail dynamics in several key ways, so you can't just blindly use highway formulas. Start with the fact that the wheels of a truck are flexible rubber acting on a surface made of slightly fluid-like asphalt, while a train has steel wheels acting on steel rails. And you have to consider things like cant deficiency. A train flying through the curves at 6 inches of cant deficiency, no matter how light, would wear the edge of the outer rail more than a heavy freight going through the curve at its balancing speed. Of course that still says nothing about the various other stresses in the rails and the whole track structure in general, but I can tell you that it's not really anything like that of rubber on asphalt at all.

    ReplyDelete
  36. Adirondacker31 May, 2009 21:55

    I'm not an expert, but I do know that track wear is proportional to the fourth power of axle load (link).

    But your link is to a discussion of road wear. With a table showing the differences in wear between "flexible" road surfaces and "rigid" road surfaces. How flexible is a well maintain railbed compared to a highway? A link to a document with this caveat "Therefore, as a rule-of-thumb, the damage caused by a particular load is roughly related to the load by a power of four (for reasonably strong pavement surfaces)." They then give an example using the more accurate formula for flexible surfaces and the results using "4th power"

    When it comes to axle loading they may be very very similar. But the things that wear highways aren't necessarily the same things that wear railways.

    How much does it cost to maintain a separate fifth track versus the added maintenance of running freight on one of the local tracks?

    For my amusement I wandered in to the DEIS for the Cross Harbor Freight Tunnel. They are hoping to divert enough traffic to run 68 trains a day, enough to improve some air quality standards by 5 percent in metro NYC. Things about expanding yards in Maspeth and building new yards at the site of Pilgrim State Hospital. Passenger service during temporary emergencies. Nothing I could find, in my quick reading, about building significant stretches of new track on the Island. Where the LIRR, the country's busiest railroad operates. Hmmm.

    ReplyDelete
  37. The LIRR runs FRA-compatible trains, and a couple of freight trains via Hell's Gate Bridge. If they ran heavy freight on the subway, then you might have a case.

    ReplyDelete
  38. "For my amusement I wandered in to the DEIS for the Cross Harbor Freight Tunnel. They are hoping to divert enough traffic to run 68 trains a day, enough to improve some air quality standards by 5 percent in metro NYC. Things about expanding yards in Maspeth and building new yards at the site of Pilgrim State Hospital. Passenger service during temporary emergencies. Nothing I could find, in my quick reading, about building significant stretches of new track on the Island. Where the LIRR, the country's busiest railroad operates. Hmmm."

    Yes, well, they're planning to use the essentially unused LIRR branch from Brooklyn to Maspeth. Why would they need new track for that? This also connects directly to the freight-only link to the Hell Gate Bridge; no interaction with the major passenger lines of the LIRR is needed.

    As for the trains continuing deeper into the Island, it would be a small minority of trains and would presumably run at night. The idea is to pull big trucks off the Hudson River and East River crossings (and to a lesser extent the Bronx River crossings) and out of Manhattan (and to a lesser extent Brooklyn, Queens, and the Bronx). Actual freight service to Long Island beyond Brooklyn and Queens is only a minor factor.

    ReplyDelete
  39. And the Hell Gate line goes where? To the Hudson and New Haven lines of Metro North, the country's second-busiest commuter railroad. The New Haven Line is also part of the NEC and has Acela and Northeast Regional trains. And while the LIRR no longer operates freight, there is indeed a freight rail service on the Island, operated by the New York & Atlantic, which runs on LIRR tracks, often during the day.

    ReplyDelete
  40. Yes, and the New Haven Line runs FRA-compatible trains as well...

    ReplyDelete
  41. As for the trains continuing deeper into the Island, it would be a small minority of trains and would presumably run at night.

    Which must be news to the NY and Atlantic. YouTube is full of videos of long freight trains running through stations or grade crossings farther out on the Island than Maspeth. During the day.

    Pilgirm State Hospital, what's left of it, spreads out all over. I'm assuming, but I may be wrong, that they would be building something near the Deer Park station, well past Maspeth, all the way through Nassau County and into Suffolk. They wouldn't be considering building a yard unless they had traffic to make economic.

    The idea is to pull big trucks off the Hudson River and East River crossings.

    And off of the Long Island Expressway, the only limited access highway that carries commercial traffic into Nassau and Suffolk.

    There's millions of people out on the Island. They eat. They drink. They use drywall and lumber. They get UPS packages delivered - there's a video of a hundred car train passing by with at least 5 UPS trailers on the train. They're going to be doing more than dumping freight in Maspeth.

    The LIRR runs FRA-compatible trains...and the New Haven Line runs FRA-compatible trains as well .

    What does that have to do with how freight trains wear track? Or have you decided to change the argument from freight trains wearing track heavily to not mixing FRA compliant trains and non compliant trains?

    ReplyDelete
  42. My point is that since these lines have compliant (i.e. heavy) passenger trains, they have to be maintained based on heavy axle loads regardless of how much freight runs on them.

    ReplyDelete
  43. Alon, FRA compatible trains may be "heavy" compared to their non-FRA-compatible counterparts, but 120,000 lbs is still not a 268,000lb freight car, or for that matter even an 185,000lb (non-FRA-compliant) electric locomotive. I don't think the freight is as much of an issue for track as you think, especially given that in Europe both CTRL/HS1 and the Perpignan-Figueres line are designed for shared use by full speed TGVs and freight. Anyhow, on the Peninsula, I think the only way you'd save money is by maintaining one pair of tracks to a lower standard, which means lower speed, which means less operational flexibility in case of delays, track closures, etc.

    ReplyDelete
  44. I thought European freight trains were lighter than American ones, and the ones running on HSR tracks even lighter. Am I wrong?

    ReplyDelete
  45. @Alon, yes, the typical Euro axle load limit is 22500 kg, vs. 32000 kg in the US (i.e. US trains are 40% heavier per axle)

    ReplyDelete