Wednesday, December 3, 2008

Going In Depth on the Peninsula

NOTE: We've moved! Visit us at the California High Speed Rail Blog.

You may have noticed a new addition to the blogroll at right - the Caltrain HSR Compatibility Blog. It's a fantastic site run by Clem, who lives on the Peninsula and is an expert on the Caltrain (and now HSR corridor). His blog provides some very detailed discussions of the HSR route between SF and San José and focuses on what needs to be done to build HSR the right way.

His posts offer some of the most informative commentary on the Peninsula segment that I've seen. For example, his San Bruno post pointed out a significant issue that will have to be addressed to make HSR trains run effectively:

One important consequence of this history is that San Bruno is left with one of the sharpest curves on the peninsula, where the cutoff formerly diverged from the old main line. The radius of the curve is 1800 feet (550 meters), giving it a maximum safe speed of about 70 mph (115 km/h). Caltrain has a speed limit of 60 mph at this location. Slowing down a high speed train from a peninsula cruise speed of 125 mph (200 km/h) to take the existing San Bruno curve would cost more than a minute, or over half a percent of the entire SF to LA running time. Considering how much investment is being made to shave seconds off run times for the entire system, a 1-minute penalty in San Bruno for a single curve should raise some red flags at the CHSRA.

Consider the "San Bruno Curve" a significant obstacle to HSR on the peninsula.

His Millbrae post from yesterday offers some excellent thoughts on how the flaws of the BART project mean that the Millbrae station is going to have to be significantly reengineered for HSR and Caltrain to be effective:

Plans for the Millbrae HSR station, described in the CHSRA's environmental impact documents, call for four tracks to pass through, with two outside platforms.

Unfortunately, the incompetent design of the existing Millbrae station (opened in 2003 and billed as the largest multi-model station west of the Mississippi) has squandered much of the space available within the Caltrain right of way. Massive support columns for the expansive station mezzanine have been placed right where you'd want to run extra tracks, and a third platform track at the south end of the station (presumptuously built to terminate some Caltrain services, assuming everyone would want to ride BART into San Francisco) sits unused, with no possibility of continuing north through the station. The photo at right, taken in 2000 during construction, illustrates how a forest of concrete pillars now constrains the right of way.

The total lack of foresight in this design is breathtaking.

To make room for four tracks, the existing southbound platform and part of the mezzanine that it supports (including the ill-placed pillars) will have to be demolished and rebuilt a few dozen feet further to the west.

Clem is determined to make sure that HSR is built the right way along the Peninsula:

Let's finish on a positive note: to avoid becoming the boondoggle that proposition 1A opponents fear, the HSR project needs to be designed under the watchful eye of an independent panel of experts with local knowledge, and especially some TEETH to bite back at the agencies and contractors as required to keep them attuned to the public interest. Whether the provisions attached to Proposition 1A will provide for this remains to be seen.

I fully agree, and that's where these HSR blogs become so valuable. We have a lot of people here who are supportive of the concept and who understand the technical issues involved. WE can provide the oversight and public pressure to ensure that HSR becomes the success we all know it can and must be for our state.

When I started this blog back in March it was primarily to fill a void - there were a few sites that had HSR content but they weren't being updated often, and weren't providing the ongoing information and insight that was needed to ensure that Prop 1A would pass. I like to think this blog filled that need and will continue to do so over the next months and years. But if we're going to build this the right way, we need local writers and activists to get involved. Kudos to Clem for stepping up on the Peninsula.


Rafael said...

One factor greatly complicating matters is that FRA has not yet ruled on whether Caltrain will be allowed to operate mixed traffic, i.e. new UIC-compliant EMUs alongside legacy FRA-compliant diesel locomotives and unpowered cars.

Electrification of the SJ-Gilroy section would likely happen in a later phase, especialy since Santa Clara county hasn't committed to paying for any of it so far (SF and San Mateo counties have).

If Caltrain gets a waiver, it's quite possible CHSRA will also get one for sharing track with Caltrain on the few short sections where adding brand-new tracks might prove very difficult: Santa Clara Alviso line turn-off, Mountain View I-85/station, Millbrae-San Bruno, tunnel near 101/I-280 interchange in SF, tunnel at 22nd Street in SF, DTX tunnel, perhaps others.

It's pretty important that FRA assign a senior liasion officer to handle all of the rail projects in California, especially wrt mixed traffic issues.

SMART would also love to use UIC-compliant DMUs but will have share the line with NWP's mile-long and FRA-compliant heavy freight trains bearing ore.

Metrolink has caused one disastrous and one embarassing accident in recent weeks, clearly there is a need for some active safety measures (including but not limited to PTC) to prevent train-on-train collisions.

Basically, the incoming administration should view California as a testbed for drafting new rules that enable cost-effective passenger service operations without breaking the freight operators' business model. It's not an easy circle to square, primarily because FRA has historically acted as a champion of the rail freight industry.

Clem said...

Rafael, Caltrain has indicated that the waiver is 90% certain to be granted, regardless of HSR.

With HSR in the mix, the peninsula will have advanced cab signaling with PTC, the equivalent of ETCS level 2 or better.

Slap that on to the Union Pacific trains used on the peninsula, and the safety case will be a slam dunk.

The Metrolink crash was an aberration of the 19th century systems and procedures. It would never have occurred with a modern (post-1980) signaling system.

randy said...

For the extension to the Transbay terminal, has anyone considered working with the mostly funded SFMuni Central subway. It goes directly from the Current caltrains station north, It would seem that logistically, adding a second layer for HSP/Caltrains would save enormously on the expense. It would get us halfway to the Station.

The connection might even save some money in that the spoils(dirt) could be hauled out on the caltrains lines?

nikko pigman said...

Well more important than just shaving a minute off of the time, it's a waste of energy (particularly for HSR trains which are meant to run at constant speed) to decelerate and accelerate again after leaving the curve. That's where the real catch comes in. Of course, I'm not sure how much energy is actually lost and how much money would be wasted and whether realigning the curve would have a good cost/benefit ratio.

yeson1a said...

Its a great site..and we need people like him! Lately there is some hyper woman posting and making up all these HSR horror stories in the local on line papers. She says it going to destory Palo Alto and Menlo Park and I think ive seen him debunk her. This is her 2nd try as she was very anti-prop1A

mike said...

nikko - Actually it's really the lost time that is most valuable. HSR will use dynamic braking, i.e. it will turn the traction motors in reverse and use them as generators to feed electricity back into the catenary. Sort of like a hybrid, but better. So braking/accelerating does not waste as much energy as you would think. And even if it did, the time lost would still be more valuable than the energy.

Andrew said...

I think the Metrolink crash in Chatsworth really highlights the need for multi-tracking along the entire LOSSAN corridor. Yes, PTC is needed now, but more capacity is also needed in the near future.

Due to being a Santa Barbara native (though currently living in Japan), and the fact that Santa Barbara County voted yes on 1A, I really feel that frequent Pacific Surfliner service taking around 90 minutes to LA Union Station is in order.

Alex said...

I have a bunch of questions that I was wondering if anyone could answer for me, this seemed like the appropriate time. Some might be stupid but I've been wondering about them since I became obsessed with the HSR project.
1) So from what I can see from the extremely informative Compatibility Blog, the plan is for 4 rails wide from San Francisco down to at least SJ Diridon, correct? I've been confused as to how the HSR train will actually get out of the bay area and onto its own dedicated track through the central valley. Looking at the pictures on the site, the HSR train will share the center two tracks with Caltrans Express lines and the outer two tracks will be for normal Caltrans service and freight.
2) Will it become a problem sharing a set of tracks with Caltrans or will there be scheduling to ensure that HSR has a straight, unhindered shot down to SJ?
3) Looking at the example of Millbrae, will all stations along the HSR line require 4 tracks to allow for full speed express trains traveling through the station and local trains or can it work with only 2 tracks?
4) If HSR shares its tracks, will the heavier Caltrans trains damage the tracks at all and require the HST to slow as the years progress? 5) Will all 4 tracks be upgraded during the Caltrans electrification process to allow the HSR to run on them or are they already ok for the speeds the train will be doing through most of the bay area (~125 mph)?
6) Which train setup do you think Caltrans will adopt after electrification when it is finished in around 2014 (is this the expected completion year?), EMUs or simply upgrading the locomotives to electric units and keeping the same passenger cars?

Thanks for all of the help. I'm in college right now but see the completion of the HSR project and its parallel projects to effect my future significantly and this blog has been a great read everyday and provided a lot of information.

Rafael said...

@ clem -

that's precisely the sort of thing I had in mind when I wrote up the rapid rail post over the weekend.

@ randy -

I don't think it's quite that simple. The central subway line is still above ground as it passes 4th & King because of the Mission Creek outfall. It needs to descend underground somewhere near Howard, hopefully they'll be smart enough to implement a level crossing the existing Muni subway line down Market. Otherwise, the Central Subway line would have to dive under BART.

HSR trains need significant curve radii, e.g. 600-800 ft minimum. That might be hard to implement at 4th & Townsend.

Personally, I'd just as soon see the TTC approached via the median lanes of 101 from Brisbane up to avoid that $3 billion tunnel altogether. That would put the train station at the TTC on the 2nd floor and the buses underground. It would also give the TTC train station a loop track for greater throughput.

The portion above Clementina could be used to run through a second building used as a high speed cargo rail yard, with last mile distribution trucks/vans at grade level.

Best of all, approaching via 101 north would create the option of tracks across the Bay Bridge, assuming it is strong enough to take the weight of the trains.

Unfortunately, sacrificing existing freeway/bridge traffic lanes is a major no-no as long as the car is still king.

@ Andrew -

please take a gander at this map. Note in particular the Metrolink lines from LA US down to Long Beach and the one across to Anaheim via a tunnel under Disneyland. A loop east of Anaheim ARTIC would allow the Disneyland line to return to LA US.

Note that I've included two options for moving the alignment away from the beachfront at San Clemente so it can be double-tracked and electrified: one follows the existing tracks to San Juan Capistrano and then hops onto the I-5 median through Torrey Pines. The other stays on I-5 to reach the upper level of Anheim ARTIC before continuing on along the 57 median to Pomona, Ontario and San Bernardino, with an optional new alignment to Victorville if the Desert XPress guys build the rest of the line to Las Vegas.

This second alignment would permit through service from San Diego to LA US as well, via a descent ramp parallel to the Santa Ana river and the aforementioned loop. There, the operator would have to choose either the Fullerton alignment or else the one via Disneyland and Paramount.

A third option is to run trains down to San Diego from Corona via the I-15 and 163 medians, terminating in Balboa Park. I think either of the I-5 routes makes more sense, but my objective was to create options other than the one CHSRA came up with. Only one set of new tracks should be constructed down to San Diego.

ladyk said...

Since there will be a lot of bottlenecks and points of contention along the line from SF to Anaheim/SD, why not put all of those into a Wiki (with some of you guys as editors) where we would have all building options, pros&cons, costs, stakeholders, current state of affairs, etc. to look up. Otherwise the information from the discussions may get lost. With all that information in one place effective oversight would be a lot easier and we wouldn't have to recapitulate the material in blogposts/comments again and again.

Andrew said...


That's interesting and probably more important in terms of ridership, but I'm more interested in the other half of the (misleadingly-named) LOSSAN corridor, the part that runs northwest of LA to San Luis Obispo, along which the Chatsworth crash happened. I know you've covered it before in your rail maps. ;-)

Robert Cruickshank said...

ladyk, that is a fantastic idea.

Eric said...


That person/parent is unreal. She has earplugs in and keeps screaming, being completely blind to her surroundings. Making up stories is an understatement!!

Anonymous said...

Running in the median of 101 won't work for reasons beyond alignment and taking highway lanes. The clearances for HSR are wider than a freeway lane, so you'd need more than the two center lanes. Second, the vertical clearances for overpasses over highways is lower than that for rail. There are also sharp curves along 101 in that section. The grade between Harney Way and Third Street is probably greater than 4%. Also, take a look at the 101-280 interchange. There and at the 101-80 interchange and at Fifth Street there are left exits from the freeway.

You would need to remove buildings to get from 80 to the TTT, and the grade down into the TTT again would probably exceed 4%. If you are following the existing off-ramp, the radii into the terminal is too tight and you will not be able to get 400m' platforms in that configuration. Finally, the highway structures are not strong enough to carry rail.

Spokker said...

Some of your guys' dedication to HSR in California is a sight to behold. But is this a form of charity to you? You guys have practically designed the whole goddamn system on these blogs with zero compensation. I wonder if Parsons and Brickerhoff or whatever the hell they call themselves isn't lurking on here and taking notes.

I'm thinking that a guy like Rafael should be hired to the CHSRA. Robert Cruickshank has done an admirable job promoting the project. Where's his cut? He deserves it.

Rafael said...

@ Andrew -

the section between LA and SLO is part of the Central Coast corridor, see page 2 of this map for details of my suggested rapid rail route inland of Vandenburg AFB.

It looks like the Chatsworth tunnel is single track, that would have to be double tracked to speed things along. Essentially the same alignment but immediately north of the current track. If preferred wrt elevation profile, the second track could be a single, longer tunnel.

Vandenberg AFB is also a single track stretch and apparently, so poorly maintained that FRA has had to severely limit top speed there.

k;illmelopark said...

@spoker ..I miss you fuc!*onk with the Menlopark/arethon/paltoalto/NIMBYS whiners

Spokker said...

Haha, thanks. I don't think I'm f'ing with them though. I'm just giving it back to them as good as they dish it out.

What I find odd is that Google News shows posts from Palo Alto Online's forums, not real news articles. Just a quirky Google thing I guess. But that's how I find those awful discussions.

Anonymous said...

@ Kill em spoker ..Glen

Warren said...

I add to Spokker's comments about the dedication of those here. Nice to see the enthusiasm and constructive comments. May be worth catching up with some of the guys at GreenGauge21, another non-profit, HSR promoter/proactive group here in the UK.

A few of the questions around mixed use of CalTrans/HSR on the Penninsula have addressed the heft of the trains and speed differentials. Provided that the rails are dedicated passenger service...there is little problem of however many trains run on it.

I do offer that mixed use of commuter diesel, commuter electric, and high speed commuter electric is used throughout Germany and UK, and to some limited use in France. Anyone who has been on the (excessively expensive) Heathrow Express will have no doubt seen the diesel service from Paddington to Bristol run along-side them going into/out of London. The main Western Mainline is not electrified (thanks largely to Brunnel's small tunnels).

While the Euro trainsets may be lighter than the American, I can't imagine a 1/4 mile long train's overall weight on track would be that different from the relatively short CalTran trains. Thus, negating any concerns for rail deformation.

A further, and interesting testbed for mixed long distance and targeted commuter service on HSR, will be the eventual use of the HighSpeed 1 (Channel Tunnel to London) by SouthEastern Trains high speed commuter service using Hitachi trainsets. This certainly could be an interesting model for a Gilroy-San Jose-San Fran high speed commuter service with CalTrans...especially if the Penninsula commuter route is the first to be upgraded to full electrification/grade separation.

Rafael said...

@ spokker -

I think you may be overstating the value of my contributions here. It's one thing to come up with some ideas for resolving thorny details, but actually executing on that (e.g. kicking BART out of Millbrae) is 99% of the work. Ok, maybe 98%.

@ Alex, Warren -

wear and tear on the trackbed (principally, geometry creep) is function of axle load, not total train weight.

In a traditional train configuration, the highest axle loads are generated by the locomotive. FRA regulations on mixed traffic mean that passenger trains in the US must use locomotives that are much heavier than needed for passenger service, so they cause more wear and tear.

EMU passenger train conforming to UIC standards come in at 14-17 tons (metric) per axle, which is very light in comparison to heavy freight trains with double-stacked flatbed cars. Spreading the traction motors out over the length of the train makes a big difference. Axle loads are also critical for achieving high speeds through corners.

The Caltrain corridor is mostly wide enough for four tracks side-by-side (i.e. 100 feet), but some sections are narrower. CHSRA intends to acquire the additional land needed, by eminent domain if need be. An alternative is to stack the HSR tracks on top of Caltrain's using an aerial structure, but that means only HSR gets grade separation in that stretch. Doable, especially with quiet zone measures to avoid having to activate the horn, but not what the current plan calls for.

Alex, you may want to consult the Caltrain 2025 docs wrt to the equipment they intend to use after electrification. As for getting the center pair of tracks to veer off, that requires either a tunnel or an aerial structure. For details, look at this satellite image of the ramp between Millbrae and SFO.

James said...

Posted on CLems blog:
(point of order, should cross-posting be kept to a minimum? Is this cross-post unnecessary?)

Would it at all make sense to run the peninsula HSR from Santa Clara, up along Lafayette St. towards Alviso and then along the berms of the old salt ponds and on viaducts along the edge of the bay until it can rejoin the ROW next to the Dumbarton Bridge and through East Palo Alto and East Menlo Park to a station in Redwood City. If this can be done with minimal environmental impact to the bay, it would bypass some difficult parts of the Caltrain ROW, such as Mountain View, and would stay away from many NIMBY problems. It could also split at Alviso with a branch to Oakland.

Rafael said...

@ James -

interesting idea, but any new construction in the salt marshes would run into serious opposition from environmentalists because of the Don Edwards National Wildlife Refuge.

Also, most of the remaining UPRR ROW in the stretch between Santa Clara and Alviso is extremely narrow, just wide enough for a single heavy freight track. You couldn't stack two HSR tracks on top of that side-by-side even if you wanted to.

Finally, the western trestle of the Dumbarton rail bridge burned down in a 1998 fire that the local fire marshall called "suspicious". Somehow, I doubt the residents of tony Menlo Park and Atherton would be any happier about HSR if it leveraged the ROW from the Dumbarton Bridge to Redwood City than if they just ran along the Caltrain ROW. Some of them are going to be implacable NIMBYs no matter what.

Casey said...

Clem's blog outlines a whole lotta of headache and cost on the Peninsula caltrain leg. Would it be more feasible and cheaper (and avoid alot of NIMBY generated headache) to put the SF-SJ section down the 101 corridor, (elevated, down the center). Jeez, cut Caltrain ball and chain loose, and just build HSR right. Seems like a whole lotta grade separations at $20M a pop, short sited overpasses, historic shacks, and whining NIMBYs (and and hell of a cost), that would be avoided. Caltrain ROW seems like more trouble than its worth.