16 February 2009

Port Pork

The SF Examiner had this interesting story on Sunday, about the Port of San Francisco's desire to grow its revenue by using HSR funds to increase the clearances inside the tunnels into the city:
Port officials hope high-speed rail money can help with tunnel needs

Importing cars through San Francisco could earn $2.5 million a year for the Port, but the Caltrain tunnels through which car-carrying freight trains would need to pass are 6 inches too short
That's $2.5 million, with an 'm', per year... for an agency that has earned a grand total of $2.6 million since 2001 from the import of bulk commodities. That's right, a piddly $400k per year. Now they propose the following:
Expanding those tunnels could cost tens of millions of dollars (...) Some of the rail bonds could be used to expand the tunnels or bore new ones.
Even if they did get those coveted six inches of tunnel clearance to run excess-height autorack freight cars, the tunnels will eventually be electrified with overhead wires, which require significant additional clearances. Let's imagine the costs associated with that:
  • The cost of increasing the ceiling height of the tunnel itself, including high-voltage overhead contact system (OCS) clearances, in order to accommodate excess-height freight cars. This cost is acknowledged at several tens of millions of dollars
  • The cost of increasing the vertical clearance of the OCS along the entire peninsula corridor to Santa Clara (where freight trains move off the corridor), to accommodate excess-height freight cars
  • The cost of procuring electric rolling stock for HSR and Caltrain with non-standard pantographs that can reach to heights of 7 meters, a value unheard of in off-the-shelf European or Asian designs
The cheapest option would be to cut the Port a $25M check so that they go away forever. Better yet: would somebody please put this agency out of its misery?

UPDATE: a useful cross section drawing of the tunnel in question, and the various loading gauge dimensions (the relevant clearance for autoracks is AAR Plate H).


  1. Actually, it's none of those costs you mention. The plan, for as far back as I can remember, has been to build a gauntlet track through the existing tunnels, right in the center where the tunnels have the most clearance, and where the catenary which will eventually be installed won't get in the way either. That's an entirely reasonable thing to do, and wouldn't cost all that much, maybe not even $25 million. Oh, and the plan has ALWAYS been to allow Plate H clearances as much as possible (though not necessarily with the catenary energized). And pantograph reach is NOT a problem. SEPTA's MUs run just fine under wire high enough for doublestack cars. High reach pantographs are not a problem, and it's even possible to adjust them on the fly for different wire heights, as the Eurostar does when going through the Channel Tunnel.

  2. @ Clem -

    OCS clearances in tunnels can be minimized by using overhead conductor rails instead, which are ok for speeds up to 230km/h. And as Arcady points out, high catenaries are perfectly possible. In addition to the Channel Tunnel, the new Betuwe line in the Netherlands was constructed such that elevating the OCS can be done at a later date at minimal cost, if dual-stacked freight trains between Rotterdam and the Ruhr Valley become necessary.

    What I'm much more worried about is that the Port of SF may harbor ambitions of substantially expanding freight rail service in the future, messing with Caltrain and perhaps even HSR operations. I completely agree that the projected revenue doesn't justify the incremental cost and headaches. It would be better better to just shut the port down altogether. Proceeds from land sales alone would probable amount a minor fortune.

    Eliminating heavy freight from the SF-Santa Clara corridor might also allow Caltrain and CHSRA to hash out a rapid rail timetable that makes do with just two tracks in many sections, including e.g. Mountain View station, where one of the VTA light rail lines terminates.

    Of course, high speed cargo would still be possible, especially if such trains are run off-peak or the trainsets are coupled to non-express HSR passenger trainsets to avoid increasing the tph. The transshipment terminal could be co-located with the overflow yard in Brisbane. Normally, passenger trains are verboten in freight yards, but high speed cargo is arguably a very different animal.

    Eliminating heavy freight from the SF peninsula might also prompt UPRR to make its single-track Milpitas line through San Jose available for two purposes:

    1) diesel-based Caltrain service between SF and an intermodal station with BART at Union City, via trackage rights through the industrial terrain at Fremont Shinn (near Van Euw Common). This would be cheaper than a new turnout across a pond just east of there. It would also immediately relieve congestion on I-880, hwy 237 and hwy 84 (Dumbarton bridge), long before the BART extension is completed or the Dumbarton rail bridge can be rebuilt. VTA might want to add a new light rail service between Alum Rock and Mountain View on existing rails.

    Commuters may not much like having to transfer at Union City and/or Milpitas Great Mall, but in this economy, it may beat having to own an additional car just to get to work and scrambling to keep up with mortgage payments.

    2) a new diesel-based ACE service to SF 4th & King (subject to trackage rights from Caltrain and compatibility with future platform height). That would immediately ease congestion on I-580, I-92 (San Mateo Bridge) and I-680 south of Pleasanton. Commuting by train and either connecting transit or a folding electric bicycle is cheaper than owning a second (or third) car.

    Unfortunately, Caltrain's to-do list is already pretty long right now. In addition to Caltrain 2025 incl. electrification and now HSR, Montery county (TAMC) is raising funds and negotiating trackage rights with UPRR so Caltrain service can be extended down to Salinas from 2011 onward. If that happens, San Beneto county may follow suit and ask for Caltrain service to Hollister - especially now that HSR is coming to Gilroy.

  3. Rafael: Even if freight traffic goes up by a factor of two or three, that's still a half dozen trains a day, which can fit comfortably in the nighttime windows, although people along the line might object. In a hypothetical FRA-free world, I think the scheduling problem isn't intractable, although the 50 mph speed for freights is way too slow for expresses, and a bit too fast for the locals, but it might be possible to fit in freights during the midday and evening, when there are fewer trains running.

  4. While excess height freight cars can be accommodated by any number of technical contortions, each of which sounds feasible, everything about this violates the KISS principle. Keep It Simple Stupid. If the CHSRA caves in to each and every little request, the project will die by a thousand cuts. All this nonsense for a couple of million a year. Maybe.

    Technically, a gauntlet track is not so easy as it sounds. The Bayshore curve just south of tunnel #4 is supposed to operate at 125 mph (if you believe the HSR run time simulations), which leaves zero (negative!) run length to return the tracks to zero superelevation before the gauntlet turnouts and the tunnel portal.

    I say, let the mighty Port of San Francisco grow their business a bit more before graduating from Plate C to Plate H.

  5. Clem: keeping it simple, in this case, would be to just electrify Caltrain and have HSRA buy schedule slots, maybe pay for some specific improvements to increase the total number of slots. And the plans for electrification, as far back as 1992, have always included a gauntlet track through the tunnel.

  6. @ arcady, clem -

    Caltrain's electrification plans were based on operations at no more than 90mph, with an option to upgrade the catenaries to 125mph in the future where the tracks would permit it. As Clem has explained, that basically boils down to adjusting the tension on the wires.

    In that context, a gauntlet track to accommodate flatbed cars carrying two containers was feasible.

    Now, HSR has entered the picture (before Nov 4, it was basically a paper tiger). CHSRA intends to bore additional tunnels to accommodate its own tracks, with whatever superelevation they need and without any turnoffs to worry about. Hence the objective of running at 125mph.

    However, I'm not sure where CHSRA was thinking of putting these additional bores in the specific case of the tunnel #4. If it assumed that they would be to one or either side of the existing ones, there would be no conflict. However, if they were thinking of building at least one of them in the gauntlet position, there would be a conflict.

    However, plans that Caltrain made long ago - long before HSR was a factor - called for a gauntlet track for freight are subject to change. The HSR start line is a $33 billion project, with several billion more going into the related SF Transbay Terminal, featuring a trainbox and access tunnel that will be shared by Caltrain and HSR. The starter line is (perhaps optimistically) expected to run an annual operating surplus of about $1 billion after several years of ramping up ridership numbers.

    The Port of SF is talking about incremental revenue opportunity of $2.5 million a year. That's a hair on the tail that's trying to wag the dog. It's much, much more important to get HSR and Caltrain electrification right than it is to keep serving the by-now minor-league port of SF. If there's a need to off-load cars in the Bay Area, why not do it at Richmond? That port already off-loads cars at a dock on its southern tip (end of Canal Blvd).

    If it needs overflow parking capacity, how about Shelby, across the Carquinez Strait from Mare Island? That site was the location of a huge smelter during the Gold Rush days and was so contaminated with heavy metals that a federal judge ordered it covered with a very thick layer of asphalt to prevent rainwater from leaching the stuff out into the bay. I can't imagine it would be expensive to lease, because it can't be used for anything other than a parking lot and, it's right next to the main rail line.

    All you'd need is a shuttle train to take the cars up there. It wouldn't even have to be multi-level, just plenty long.

  7. @ Clem, arcady -

    just out of curiosity: how much clearance do the current tunnels have and how much more would the proposed one for the gauntlet track provide?

    If the delta is small enough (inches rather than feet), could the problem be solved by lowering the trackbed in the tunnels a little bit?

    How much of a gradient can a freight train loaded with cars stacked three high handle?

  8. @arcady: I didn't know the gauntlet track had been planned by Caltrain already. If HSR plans to straighten that curve (as implied by their run times, and as infinitely simple to do!) it might be hard to implement.

    Do you know of any other low-clearance overpasses that might prevent plate H plus electrification?

    @rafael: you said exactly the right words: Tail wagging the dog!

    Yes, I think HSR plans two bores on either side of the existing tunnel #4. That appears to be the plan for #3; the Oakdale station plans already show this configuration.

    Quite aside from that, new tunnels may not in fact be necessary for a while after HSR service begins...

  9. The problem with the existing tunnels is their arched roof. It's actually quite common for railroads to take tunnels like that, and reduce the double track to a single track directly in the middle. The difference in clearance is huge, but of course the capacity is reduced. The gauntlet track allows them to have it both ways, to some extent. Also, I really doubt that 125 mph will be allowed through the existing tunnels, especially Tunnel 4, partly because of the effects of two trains passing each other at 125 in such a narrow tunnel, and partly because the tunnel actually curves at the south end, and the relatively tight clearances limit superelevation.
    As for the rest of the line, all the bridges are tall enough to allow physical clearance for Plate H cars under the wire, but a few aren't tall enough for electrical clearance, so any high freights will have to be run at night.
    I'm actually a little surprised that HSR is planning two separate tunnels on either side of the Caltrain line. I kind of assumed they'd just build a single tunnel in the ROW space that was reserved for the extra two tracks, on the west side of the line (and where there's actually already a parallel Tunnel 2).

  10. The port has a good point. They need to work out a shared cost deal with hsr and caltrain to properly upgrade that stretch of the corridor for future growth and benefit for all. Now would be the time to get it done. New business for the port is good for sf and would bring the kind of good jobs we need here.

  11. Pitting rail agencies, or any public agencies against each other is not acceptable. Now is the time to make the right investment choices to benefit everyone involved., not one over the other. This is the kind of thing the holds up progress. Close the port of sf? no way, you are pitting railroad jobs against port jobs... you can't do that. They are on the same side. Cooperation and good future planning are what's needed. Make a deal and work it out. I fully support high speed rail but not to the detriment of others.

  12. @ Jim -

    that's where high speed cargo comes in. Similar jobs in transshipment, higher-value freight.

    The SF facility could be at the Brisbane yard. A separate one could be another one near 82/E Alma in south San Jose. There's an existing turnoff just south of 87/Almaden that would provide access. Just electrify, add some four-quadrant grade crossings, a turnout toward Gilroy.

    Other facilities would be at Castle Airport, Bakersfield, Palmdale Airport, LA (just north of Union Station) and Anaheim (just west of ARTIC) in phase I. Sacramento, a location in the Riverside/Moreno Valley area and San Diego (e.g. near Miramar) would complete the HSC network in phase II.

    HSC trainsets are self-propelled and usually operated at around 250km/h (150mph) top speed. They can be run stand-alone (e.g. off-peak or at night) or else driverless, by coupling them to "local" HSR passenger trains.

    That works best if the HSC yards have through tracks that allow these passenger trains to make quick stops as they detour through them. With Scharfenberger couplings, dwell times are on the order of seconds. In the US, passenger trains are normally verboten in freight rail yards. However, HSC is a brand-new animal for which different rules could be drawn up.

  13. @ arcady -

    without the aid of a map, I'm not really sure which tunnel is which, but bear with me please.

    First, I'm not clear what you mean here:

    "As for the rest of the line, all the bridges are tall enough to allow physical clearance for Plate H cars under the wire, but a few aren't tall enough for electrical clearance, so any high freights will have to be run at night."

    Clearance doesn't change with the time of day, does it? Plus, the use of overhead conductor rails rather than regular OCS wires to pass under low ceilings reduces the difference between physical and electrical clearance to literally an inch or two.

    Second, I'd encourage that HSR avoid constructing any new tunnels south of 4th & King. Any money saved should go toward JPB's and CHSRA's shares of the budget for the expensive DTX tunnel and the SFTT trainbox.

    HSR should share existing tunnels with Caltrain and UPRR via appropriate signaling and an FRA waiver/rule of special applicability. If that means running a little slower than currently planned through a tunnel with a gauntlet track, so be it.

    However, it's really worth asking if running any plate H cars at all in the SF peninsula is even appropriate. Between Caltrain and HSR, the corridor will be quite congested during the day and peninsula residents will object to any (substantial increase in) freight trains running at night.

  14. CHSRA made it absolutely, explicitly, unambigously clear that cost is no object via its $15+ billion added cost, 100% corrupt Pachecho scam, fronted unambiguously and explicitly by the "failure is its own reward" team of Quentin "BART to Millbrae" Kopp, Rod "Father of VTA Light Rail" Diridon and Steve "$5 billion Bay Bridge cost overrun" Heminger.

    If the mighty city fathers of mighty San Jose get to funnel billions of dollars straight into the pockets of their BART-to-San-Santa-Clara promoter friends at PBQD, just why shouldn't the mighty Port of San Francisco pile on a just few hundred million or two (freight to the Port of SF is just as viable and cost-justifiable a project as BART to Warm Springs, Milpitas, "downtown" San Jose and Santa Clara), and why shouldn't the line be in tunnel all the way from Gilroy to Redwood Junction, and why shouldn't CHSRA pay for a Community Childcare Anti-Terrorism Park Senior Center in Belmont, and why shouldn't it pay to build a dedicated high speed FRA freight car handling facility and access line at Pier 80 in SF?

    Fair's fair! Spread it around!

    Come on, it's not like this should be any sort of surprise to any of your who voted for Prop 1A.

    Anybody who isn't on the gravy train is a sucker. Pile on!

  15. Rafael: the tunnels are numbered consecutively from SF. Tunnel 4 is the southernmost one, right by Bayshore. You can see the curve on the ROW maps. And while clearance may not change with time of day, voltage on the wire very well might. For the bridges I mentioned, there would be enough clearance to physically fit a Plate H car without hitting the wire, but not enough to keep the wire from arcing if it's energized, and 25kV can jump about a foot. Hence the need to run the tall freights at night, with the power turned off.

  16. @Rafael -- you're right --I'd much rather see high speed freight. I wonder if they can convince the port to invest in moving their products via hsf.. That would be the most forward thinking solution.

  17. @ Richard Mlynarik -

    I always find your diatribes informative and entertaining, but is there any way you could tone down the vitriol just a little bit? Pretty please?

    @ arcady -

    ah, I thought as much. SP did always measure everything from San Francisco. As for the electric clearance, thanks for clarifying the term.

    My understanding is that that AC OCS involves segments that are powered independently of each other and may be even be out of phase. Trains briefly disconnect their motors (via the power electronics) and coast through the interface locations.

    Normally, sections are fairly long, but in this particular instance, there could be a very short special segment covering just the overhead conductor rail in that one tunnel and the approaches to it. That section could then be temporarily taken off the grid by the dispatcher, just long enough to allow a plate H train to safely pass - even during the day.

  18. @ Jim -

    High Speed Cargo is not a drop-in replacement for regular bulk freight. It's a niche freight market for highly time-sensitive high-value-per-ton goods, such as mail & packages, air cargo containers, perishable foods, fresh cut flowers etc.

    For the latest in efficient intermodal freight, check out the French Modalohr system being rolled out in Europe. An bi-level version (upper level aluminum), supported by special lift trucks at the transshipment facilities, would be an awesome basis for shipping new cars quickly and efficiently to multiple inland distribution centers. The snag is, the Modalohr system does not meet FRA buff strength standards and the axle load is too high for operation on HSR tracks (strict limit of 17 metric tons). Besides, it would probably be impossible to travel at 150+ mph, even with aluminum enclosures to improve the aerodynamics.

    A bi-level Modalohr system for shipping new cars to dealers would be entirely different from the bi-level Eurotunnel car ferry trains. Their upper deck is loaded with an elevated access road on the other side of the train (not shown in the video). There are no seats, people stay in their vehicles during the 35-minute journey unless they need to use the restrooms located at either end of each train. There are also tall single-level cars for buses and commercial vehicles. The enclosures and internal doors are for fire safety, smoking is prohibited.

  19. Rafael: I'm fairly sure that conductor rails aren't used with AC electrification except in some very, very special cases like drawbridges. The biggest users of overhead conductor rails are probably the metro systems in Spain, and it makes sense when you have a DC system with relatively high currents, where if you had wires, you'd need two contact wires and a parallel feeder. With AC systems, you don't need lots of current, and thus the overhead rail just adds weight, and probably pantograph wear.

  20. Oh and regarding sectioning, yes, the overhead is divided into electrically isolated sections which can be switched on and off individually. At crossovers and the like where speeds are low, insulators are used in the wire. At higher speeds (above 45 or so) and on mainline track, they have a section where two wires overlap but aren't electrically connected. At substations and halfway between them, you may have different phases on two adjacent sections, so there's a neutral section between them, so that the pantograph doesn't connect two different phases. The power in the train is actually tuned off not by power electronics, but by a high voltage circuit breaker. In general, I suspect that it wouldn't be a good idea to have circuit breakers in the power supply being opened and closed constantly, since it's yet another moving part that can fail. Explosively.

  21. @ arcady -

    we're still talking about 2-3 freight trains per day, aren't we?

    The high voltage circuit breakers are used because they're cheap and very rarely activated. In this particular case, you'd use a GTO instead.

    Btw, overhead conductor rails are used all around the world at speeds up to 230km/h. I don't know if any application is AC, but it's certainly not just for metro systems in Spain any longer.

  22. "Rafael: I'm fairly sure that conductor rails aren't used with AC electrification except in some very, very special cases like drawbridges."

    London's Crossrail is doing it. That is a *very* long tunnel for an overhead-electrified route, and the several-inch savings in tunnel clearance was worth it for them.

  23. "just out of curiosity: how much clearance do the current tunnels have...?"

    Just guessing from the cross-section Clem provided--

    If we assume the semicircle top of the tunnel is centered 8.5 ft above the top of the rails, and the track centerlines are 13 ft apart, then a load 20 ft 2 inches high and 8 ft wide will just contact the tunnel arch if we lower the track by 0.95 ft, or shift it toward the center of the tunnel by 1.07 ft.

  24. I forgot the arch apparently no longer has a 15-ft radius-- the shotcrete seems to be 3? inches thick inside that radius.

  25. Lets discuss the environment first. The Port of San Francisco is a natural deep water port with a natural depth of over 30 ft in most places. Most ships today require a depth of 42 ft. That being said the Port of SF only requires dredging every 2 to 3 years to get the deep draft that the modern ships require. The Port of Oakland on the other side of the bay is not a natural deep water port. It requires dredging annually to maintain just 20 ft. It of course is dredged deep enough to handle the largest ships that can go under the bridges. The point is that for rail Oakland is a better port since it is served by both the Union Pacific and the Burlington Northern Santa Fe. When it comes to the water San Francisco is the better Port. If the Port of SF was served by two Class 1 railroads it would be the winner even though trains must traverse the bay simply because the Port of SF is cheaper and better for the environment to maintain. That being said when the economy rebounds and the Port of Oakland is at capacity the Port of SF should be the next viable Port to accommodate the overflow. The other thing to consider is that Oakland is a container port not a bulk commodity port. It does not handle automobiles anymore either (except at Matson for Hawaii). The Port of San Francisco got smart by giving up the container business to Oakland and focusing on what Oakland does not handle. This allows for both ports to thrive, providing high speed rail and Caltrain don’t stop them. We need both ports for our future as a nation and for the growing population of the bay area. It would be stupid of us all to close off the Port of SF to the national rail network by only focusing on moving people by rail. It is equally important to move people by rail as it is to move freight by rail.

    Lets discuss the law: California Public Utilities Commission General Order 95…this order requires that on any rail line where there is freight the height between the top of the rail and the bottom of the wire requires 22.5 feet of clearance. This must be preserved on this line. Rail cars are getting larger in all aspects. The reality is that in the distant future 22.5 feet might not be enough. This height will accommodate any existing rail car. This is also good for the passenger train. Who is to say that passenger cars won’t get taller either?

    How about those tunnels!? Built in 1906 and sprayed with shotcrete in the 90’s. Does the shotcrete make everybody feel all warm and fuzzy. If so we are all stupid. Everybody is expecting those 1906 tunnels to accommodate 100 commuter trains a day high speed rail and freight trains. Did everybody forget about earthquakes? Reality check! This is what needs to be done to make everybody happy and keep everybody safe. Undercut the tunnels to get the port the height they need now. This can be done on the weekends at night over several months with out affecting Caltrain service. Build 4 new tunnels next to the original ones that will accommodate 2 main lines. Run all trains on the new tracks and through the new tunnels that are able to handle an earthquake of 8.0 or larger (Japan does this now). Make them tall enough and wide enough to handle any type of rail car. Rebuild the old tunnels to handle an 8.0 earthquake and accommodate all types of rail cars. I know everyone is going to say it is easier said than done but it is the right thing to do. The California High Speed Rail plans call for 4 main lines anyway between San Francisco and San Jose anyway. Just do it right!

  26. California Public Utilities Commission General Order 95…this order requires that on any rail line where there is freight the height between the top of the rail and the bottom of the wire requires 22.5 feet of clearance.

    To be pedantic, GO95 does not allow for 25 kV overhead electrification. The de facto world standard is illegal in California.

    So I wouldn't put too much stock in the CPUC's regulations: they will have to be amended, hopefully in a way that makes sense (i.e. doesn't account for olde-tyme "railroad men" walking on top of freight cars to set hand brakes)

    I also note their freight cars are assumed 15'6" tall.

  27. Clem you are right about the reasoning behind the height requirement (brakeman on the roofwalk). Currently the tallest cars on the peninsula are 17 feet tall. Autoracks are 19.5'. AutoMax and double stacks are 20'2". The San Diego trolly which hosts freight trains on the line at night is at the required level and believe it or not the VTA line that goes to Moffet Field from Mountain View is at 22.5' as well. Freight was supposed to go on this line but the platforms were build incorrectly. The point is the height requirement forces the hand so to speak to protect frieght. Rail cars are getting bigger so it is vital to the Peninsula that is unless we want everything trucked in.