25 April 2015

The Blue Doors Will Open

Blue doors open at old 8" platforms (shown here)
Yellow doors open at HSR height
There are noises that Caltrain's new EMU fleet might sport a double set of doors, to enable boarding at two different platform heights.  While this isn't necessarily ideal, it is a reasonable solution given the constraints of the problem.  And yet, the prospect of trains with double the usual number of doors elicits gasps of horror from some transit advocates and industry insiders.  Let's go through some frequently asked questions to explore the roots of this choice, and feel free to ask more in the comment section.

Q. Do you intend for every Caltrain platform to be rebuilt?

A. Yes!  Today, the number of Caltrain platforms that support level boarding is zero.  In order to achieve the short and predictable station dwell times necessary to operate the blended system with acceptable reliability, level boarding is an operational imperative for Caltrain. The most important thing to realize is that Caltrain will eventually have to rebuild every single platform system-wide -- if not to the same height as HSR, then to some other significantly greater height for level boarding than the current 8-inch standard.  Level boarding is not just an option; it is a necessary expense without which blending Caltrain and HSR will fail.  Every platform must be rebuilt no matter what; this is the premise from which the rest of the discussion must start.

Q. But why insist on compatibility with HSR platforms?  Caltrain and HSR are separate systems serving separate markets, so why is this compatibility thing such a big deal?

A. Compatibility with high-speed rail is important for two major reasons.

First and most importantly, the San Francisco Transbay Transit Center will be a system-wide bottleneck for both Caltrain and HSR, with just six platform tracks.  This cramped but critically important terminus will be even more constrained if the two operators are forced to use segregated platforms.  In a segregated world, opposing flows of arriving Caltrain and departing HSR could conflict in the station approach tracks, triggering cascading delays should even one train fall behind schedule.  With platform compatibility, any arriving train can be routed to any available platform, minimizing the domino-effect of delays.  The Transbay designers know this issue is the Achilles' heel of the entire design, which is why they are pushing Caltrain and HSR towards compatibility.  The risk of an occasional equipment failure or medical emergency causing a system-wide meltdown depends on the probability of such an incident, combined with the underlying resiliency and flexibility of the infrastructure.  A segregated Transbay design is asking for trouble when things don't quite go according to plan.

Secondly, compatibility has enormous cost advantages for sharing station infrastructure, as will be seen below.  The savings from sharing station infrastructure at just four locations along the peninsula (Transbay, Millbrae, Redwood City and San Jose) could easily exceed the combined cost of converting Caltrain to high platforms system-wide.

Compatible platforms if operated carefully will not interfere with HSR security or fare collection methods.  They are solely a means to maximize the utility and robustness of the Transbay Transit Center and to reduce the capital costs of building California's HSR system by about a billion dollars (yes, with a 'B').

Q. Why can't HSR just select a train design with low floors?

A. It's not that easy.  The CHSRA has expressed an understandable preference for service-proven designs, to draw from the widest range of suppliers worldwide.  Very-high-speed trains (VHST) capable of speeds greater than 200 mph typically do not have low floors.  Nearly all high-speed train designs from Europe, Japan and China for the past several decades have featured high floors, with  few exceptions.  To achieve level boarding as mandated by the ADA while still drawing from the greatest possible selection of vendors, high platforms are almost a necessity for California's HSR system.  The only 200+ mph train with a "low" 30-inch floor is the Talgo AVRIL prototype, still in development.  It does not have distributed traction, which will be important in California's mountainous terrain.

The three foregoing questions allow Caltrain's entire range of possible solutions to be encapsulated in one simple flow chart:

Notice that ALL the level boarding solutions require dual level boarding, at the 8" legacy platform height and at whatever new level boarding height is selected.  Dual level boarding is not an easy problem to solve and usually involves some degree of awkward and clunky mechanisms, be they deploying steps, wheelchair lifts, automatic trap doors, or double sets of doors.  Even the supposedly "simple" level boarding scenario at 25" suffers from this complexity, a fact that is either glossed over or completely misunderstood by most advocates of this solution.

Q. So who cares if Transbay is so constrained?  Can't Caltrain just terminate whatever overflow traffic doesn't fit at the 4th and King terminal?

A. Emphatically, No!  Transbay is a key destination that every Caltrain must serve, especially at rush hour.  75% of Caltrain riders are commuting to work, and there are more jobs located within a 1/2 mile radius of the Transbay Transit Center than within a 1/2 mile radius of all other Caltrain stops from 4th and King to Gilroy, combined!  This enormous concentration of jobs in the heart of San Francisco will only increase with the many new office towers going up today.  Terminating even one Caltrain short of this gold mine of ridership would be quite simply counter-productive, a waste of taxpayer money and a failure to meet obvious demand.

Q. Do you understand the enormous effort and cost to do this?

A. It is a large expense, but also a necessary expense.  The cost of raising platforms is not strongly sensitive to height: rebuilding to ~48" is only slightly more expensive than rebuilding to 25", 30", or any other level boarding height.  Rebuilding to the same boarding height as HSR creates an opportunity to have the high-speed rail project defray some of Caltrain's expense for the conversion to level boarding.

The cost of totally rebuilding every platform is generously $10 million per platform, placing Caltrain's system-wide level boarding tab at (very roughly) $10 million/platform x 2 platforms/station x 32 stations = $640 million, less than half of the cost of the modernization project. On the basis of cost per minute of trip time saved, level boarding beats electrification.

Q. Doesn't this whole issue just boil down to a San Francisco Transbay problem that should be solved at San Francisco Transbay?

A. No, this is not just a Transbay issue.  Rebuilding to the same interface as HSR also enables savings of more than a billion dollars where station infrastructure can be shared elsewhere:
  • The massive dual-level elevated station, the six-mile approach viaducts, and the "iconic bridge" in San Jose would no longer be required, with platforms shared at ground level within the existing footprint of the Diridon station.
  • Squeezing a fourth track under the Millbrae station, requiring hundreds of millions of dollars of tunneling expense, would no longer be required.
  • A Redwood City HSR station, configured to provide HSR service to the booming northern end of Silicon Valley, to enable Caltrain to make cross-platform transfers from locals to expresses, and to tie in future Dumbarton Corridor service, would have a much reduced footprint and would more easily fit in the available site.
Factoring in these infrastructure savings, the cost of converting Caltrain to ~48" is actually negative.  It would be unwise not to do it.

Q. Won't rebuilding all the platforms take years and be an operational nightmare?

A. It will take years, but it can happen with relatively little disruption.  Over the past 15 years, Caltrain has rebuilt 37 platforms from the ground up.  Caltrain has a demonstrated track record for planning, funding and executing platform reconstruction projects.  You'll be hard pressed to find anyone who remembers this causing major disruptions.  EMUs with dual height boarding would considerably simplify the logistics of rebuilding platforms, since each station could be rebuilt independently as funding becomes available and as the planning process progresses in each community.

Q. Will Caltrain end up with a mix of high- and low-level boarding platforms, perpetuating this strange dual boarding height situation forever?

A.  No.  The lower set of doors has only one temporary purpose: to enable boarding from 8" platforms during the transition.  All platforms would be rebuilt to ~48", and boarding from the lower level would ultimately be discontinued.  Any remaining 8" platforms would become an impediment to the blended system, because they would introduce longer station dwells with a significant probability of unplanned delay when boarding or alighting persons of reduced mobility.  Making a mess of the timetable will no longer be tolerable, so there will be a strong incentive to finish the job even at minor Caltrain stops to ensure the highest level of punctuality and system reliability.  This makes Caltrain very different from MUNI or some East Coast commuter railroads, where a mix of boarding heights has persisted for decades because there is no operational imperative for 100% level boarding.

Q. Until all platforms are raised, wouldn't accessibility and dwell time be worse than they are today?

A. No.  EMUs with dual sets of doors would board from 8" platforms with the same efficiency as Bombardier cars, with just two steps up from the platform into the lower level of the train.  (Note that trap door designs would not fare nearly as well in this respect!)

Q. Won't dual sets of doors cause passenger confusion and long dwells?

A. Boarding the train will be obvious, based on which doors open.  Alighting requires queuing at the correct door, which can be facilitated by color-coding of the doors, LED displays, and audible messages.  "Next stop, Menlo Park.  The blue doors will open.  (...)  Next stop, Palo Alto.  The yellow doors will open."  Blue and yellow are good contrasting colors that can be distinguished by color-blind passengers.  People aren't stupid, and should someone get confused, the different doors would be within a few steps and within sight of each other so any mistakes would not lead to significant delays.

Q. Won't all these extra doors displace seating areas and reduce Caltrain's seating capacity?

After conversion to level boarding,
blue doors are plugged and replaced with seats
A. Not necessarily.  Caltrain has estimated that dual sets of doors would displace between 78 and 188 seats per train, or roughly 15 to 25% of a train's seating capacity.  This is a temporary situation during the transition to level boarding, and can be mitigated by procuring extra-wide trains with 5-abreast seating and longer 8-car trains to preserve overall seating capacity.  The space lost to extra doors can be used by standees, who do not have many good options on today's Caltrain fleet.  When the platforms are all converted to level boarding, the lower doors can be removed and additional seating can be installed, especially if this feature is designed into the new EMUs from the outset.

In the meantime, to minimize the loss of seating capacity, it would make sense for Caltrain to make seating capacity a selection criterion in the vehicle procurement process.

Q. Wouldn't this create an accessibility problem, in terms of ADA compliance?

A. No.  Just like today, lifts or bridge plates would be required to board persons of reduced mobility from an 8" platform.  There would need to be an in-vehicle wheelchair lift to change levels inside the vehicle, to allow wheelchair users to board and alight at stations with different height platforms, or to avail themselves of an accessible bathroom on the lower level.  This is not a new technology; these off-the-shelf mechanisms are no more complicated than the exterior lifts used on Caltrain's gallery cars.  An example of such a lift can be seen in this video.

Q. Will bicyclists have to navigate interior vehicle steps, potentially while the train is moving?

A. Yes.  These steps could be made wider and shallower (greater tread depth) than anything in today's Caltrain fleet.  The three steps from a 25" lower level to a 48" mid-level floor could be fitted with wheel gutters to allow bicycles to easily roll up or down along the stairs.  This would make the steps far easier to navigate than the four steep steps up from an 8" platform into a 45" gallery car, turning the corner around a pole through a crowd of Giants fans--the scenario that Caltrain bicycle riders are forced to contend with today.  And riders commonly lift, turn and sort their bicycles by destination while the train is moving, so a bit of jostling isn't exactly a new thing for the bike crowd.

Q. Wouldn't trap doors resolve this whole situation with dual doors?

A. No.  Trap doors have numerous flaws, including one fatal flaw: they would preclude Caltrain passing high platforms at speed while still maintaining an ADA-compliant 3-inch gap when stopped at a high platform.  Trains sway from side to side when running at speed, and the alignment between the track and the platform edge isn't perfect; that means the space between the train and the platform needs to be wide enough to prevent platform strikes but narrow enough to comply with ADA rules.  There is currently no rail system in the United States that can do both: there is either a speed limit when passing platforms (e.g. BART) or the gap when stopped is greater than 3 inches (e.g. Northeast Corridor).  Satisfying both constraints (< 3" gap and 100+ mph past platforms) requires a small bridge step to extend from the train when stopped.  This sort of gap filler mechanism is unlikely to be compatible with a trap door configuration.  Trap doors have other disadvantages, such as increased dwell times while the trap door mechanism is moving, sensitivity to damage from dirt buildup and foreign objects commonly found on train floors, and too many steps up from a low platform during the transition period.

The Takeaway
  • Level boarding is not just an option; it is an operational imperative for the blended system.  The blended system will not work reliably without it.
  • The new EMU vehicles must enable Caltrain's transition to level boarding, or the chance to convert to level boarding will be lost for another 30+ years, the life span of the new train fleet.
  • It is appropriate for Caltrain to adopt the same platform height and width as HSR, in exchange for the funding to achieve the transition to level boarding.
  • Transitioning to level boarding is complicated regardless of the chosen platform height; there is no easy solution.
  • Dual doors are the path to level boarding with the fewest flaws, under the imposed constraint of high-platform HSR.

08 April 2015

Find Out Your Pole Placement

Plan showing where electrification poles
will be placed in the vicinity of the
historic El Palo Alto redwood
(see RFP Volume 3 page 1193)
Caltrain has spent over $15 million on its electrification project so far, primarily for environmental clearance and preparations for procurement.  In late February, we found out where a lot of this money went: the Request For Proposals (RFP) for the electrification project was released.

This RFP is an incredibly prescriptive document that tells prospective bidders precisely what the project should look like, down to the last bolt.  Volume 3 of the RFP (download the 2840 page, 214 MB PDF file) includes layout plans of the overhead electrification system that dictate the exact placement of every single pole foundation.  The prospective contractor is admonished that pole locations cannot be changed without first submitting a formal design variance request to Caltrain.

This procurement is being carried out as a "Design-Build" where the winning bidder will be tasked with "designing" the project, which in this case will amount to a connect-the-dots exercise to duplicate Caltrain's highly prescriptive preliminary engineering drawings into final construction-ready drawings.  What little room is left for creativity and efficiency is stifled by an onerous variance process that requires the "designer" to submit extensive paperwork to Caltrain for approval of the slightest change to the design prescribed in the contract.  One can easily imagine how the goal of Design-Build contracting, namely to reduce risk and cost by consolidating decision-making under a single entity, would be lost under the hyper-prescriptive approach that Caltrain has chosen.

The thousands of pages of the RFP highlight the cozy symbiotic relationship that exists between government agencies, their in-house consultants, and private contractors.  Without an ounce of nefarious intent on the part of any of its participants, this self-reinforcing triangle, hardly unique to Caltrain, brings together hollowed-out government agencies with rubber-stamp boards run by politicians, permanent in-house consultants whose primary motivation is to justify their existence through highly prescriptive decisions that increase scope at their whim, a profit-hungry coterie of construction companies ticking all the boxes for shareholders and labor interests, and a byzantine system of contracting regulations and reporting requirements, quite ironically intended to prevent taxpayers from being defrauded.  The results of this firmly-entrenched Transportation Industrial Complex are projects that deliver less and cost more, typically three times the going rate in other first-world countries where government agencies are centralized, smart, and employ an experienced staff of technocrats whose first interest lies in serving the public with better transit at lower cost.  What can be done about this system?  Not a whole lot.  It is the logical byproduct of our decentralized system of government and of our free markets, pursuing their respective enlightened self-interests.  These self-interests include neither low cost to the taxpayer nor excellent transit service to the user.

The recently-completed modernization of Auckland, New Zealand's commuter rail network, of quite similar technical scope, is an instructive benchmark against which to evaluate Caltrain's modernization efforts.

15 March 2015

News Roundup, March 2015

Regulations for 25 kV electrification.  After two years in the making, CPUC proceeding R1303009 appears to have produced the new rule book for stringing up high-voltage rail electrification in California, although the final document is still to be formally adopted by the Public Utilities Commission.  The final draft hammered out after numerous meetings by the high-speed rail consultants, freight railroads, utility companies and other interested parties thankfully bears little resemblance to the original draft proposed by the high-speed rail consultant, which was a mess.  While the new rules apply only to dedicated high-speed rail corridors without grade crossings or freight trains, the language is flexible and there appears to be room allowed for parties such as Caltrain and UPRR to agree on those particular items on a case-by-case basis, without resorting to an entire new CPUC rulemaking process for the peninsula rail corridor.  Most importantly, the new rules provide a clear framework that allows detailed engineering design of Caltrain's electrification project to proceed with a very low risk of future regulatory surprises.

Grade separations.   An unfortunate series of accidents and suicides have re-ignited the debate over grade separations in PAMPA (Palo Alto - Menlo Park - Atherton).  This portion of the corridor abuts some of the most expensive real estate on the peninsula and is home to the most contentious and litigious environment for local decision-making (or lack thereof, as the case may be), and it is gradually dawning on these communities that Something Must Be Done.  Grade separation is often misunderstood as an all-or-nothing proposition, when in reality it is a process that has been underway for decades.  The peninsula rail corridor is already 62% grade-separated today.  Communities up and down the peninsula are starting to talk more about finishing the remaining 38% of the job.

Quiet zones. Palo Alto recently conducted a study session on quiet zones, which would stop routine horn-blowing at grade crossings.  Curiously, the systematic use of train horns before every protected grade crossing is an American practice not usually seen in other countries, where crossing gates combined with visual and audible warnings are considered to provide a sufficient level of public safety.  But then again, those other countries don't have lawyers like ours.  The staff report contains a nice overview of the process for establishing a quiet zone; the main impediment seems to be the local community's assumption of legal liability for grade crossing collisions.

Peninsula HSR reboot.  After spending $45 million on environmental clearance and design of the original and much-reviled four-tracks-all-the-way peninsula HSR project, the California High-Speed Rail Authority shelved the plans with nothing delivered.  The Authority recently submitted a Project Update Report to the legislature, stating that a new environmental clearance for the peninsula segment is scheduled for completion in 2017, with construction complete in 2028.  [UDPATE 3/21]  Furthermore, an obscure Ridership Technical Advisory Panel Review of the California High-Speed Rail Ridership and Revenue Forecasting Process dated December 2014 includes the map graphic at right, and states that the Authority "has requested a stand-alone analysis of two individual segments of the overall system, to assess the viability of each segment operating independently for a handful of years prior to the completion of the system. (...) The Bay Area segment would require improvements to existing rail infrastructure. HSR service would be blended with existing Caltrain commuter rail."  The peninsula schedule may be pulling to the left.  To clear an EIR that fast, for a project so much more controversial than mere electrification, the process would have to be re-started very soon.  This may or may not be related to the choice of...

New Caltrain CEO.  Jim Hartnett was selected as Caltrain's new chief.  He is a consummate local political insider with no background as a transportation agency executive, other than chairing the Caltrain board of directors and vice-chairing the board of the CHSRA.  While he will no doubt be well equipped to navigate the choppy waters of inter- and extra-agency politics, he is less likely to shake up the Caltrain organization or to bring fresh outside-the-box thinking to the development of the "blended system."  While Caltrain has plenty of political and organizational problems to work on, the technical aspects of blending will determine the success of the rail corridor for decades to come.  Will Hartnett do anything about the top ten problems facing Caltrain, or will he just hold the current course?

11 January 2015

Second Thoughts in Palo Alto

The Caltrain board of directors recently certified the Final Environmental Impact Report for the electrification project, the last environmental clearance necessary to move ahead with construction.  As part of this certification, a list of unavoidable impacts (which cannot be reasonably mitigated) is issued along with a statement of "overriding considerations," basically a justification for why no mitigation is feasible.

This doesn't sit well for the Palo Alto Weekly, which already wrote about the lack of mitigation for the expected worsening of local traffic around several already-jammed intersections in the city.  The Weekly ran an editorial ("Caltrain's electrification project is pushed forward with impunity", a title later toned down to "Caltrain's Bad Judgment") that calls out Caltrain for a number of supposed failures.  Read in the context of the long-running debate over high-speed rail, however, this editorial is off the mark.

Most glaringly, the editorial blames Caltrain for a failure to study grade separations together with the electrification project: "it's long past time for Caltrain to include planning and engineering costs for the least expensive method of eliminating grade crossings: raised berms and lowered roadway undercrossings."  This demand is more than a bit disingenuous, since just five years ago Palo Alto residents were vehemently opposed to raised berms (see Palo Alto Weekly article from April 2009), which were then described by the more alarmist members of the community as a "Berlin Wall".  The heavy-handed public outreach process carried out by the CHSRA five years ago thoroughly poisoned the well, and discussions of above-grade solutions still elicit raw emotions.  Case in point: a recent grade separation study commissioned by the City from tunneling firm Hatch Mott MacDonald studied only the most expensive below-grade options while pointedly excluding a raised berm solution from its scope.  Considering this context, the Weekly would do better to call for the City of Palo Alto to study those controversial raised berms.

Rather than complain, the Palo Alto Weekly should start covering what it really takes to get grade separations built:
  • Calling on the City of Palo Alto to perform a complete study of grade separations, one that does not side-step or ignore affordable above-grade solutions
  • Building strong community support for grade separations, by educating the community not just about technical possibilities but also about costs and available funding sources
  • Prodding the BART-obsessed VTA to start paying attention to the funding needs of northern Santa Clara County
  • Scraping together a local funding contribution as Berkeley once did for their preferred BART configuration
  • Working through the Public Utilities Commission, the agency that regulates all grade crossings and grade separations in California, to obtain Federal and State funding contributions
  • Applying to get Palo Alto crossings onto the CPUC's grade separation priority list
  • Undoing years of misinformation and community resentment lingering from the controversial HSR process
Grade separation of the peninsula corridor is a decadal undertaking that will eventually run its course.  The Palo Alto Weekly editorial board can either help or hinder this necessary progress, and blaming Caltrain doesn't help.

01 December 2014

Metrolink Scorns Electrification

Just like the peninsula corridor, the other end of California's planned high-speed rail network, in the Los Angeles basin, could also benefit from European-style "blended" service where electrified commuter trains and high-speed trains share tracks and stations.  Many of the solutions being developed by Caltrain for the peninsula corridor could also prove useful in the LA region, something that is not lost on rail supporters.  Paul Dyson, president of the advocacy group RailPac, recently wrote a letter to the relevant authorities expressing support for the idea of electrifying portions of the Metrolink commuter rail network to better integrate with high-speed rail.  His proposal is aptly named "Electrolink".

While the high-speed rail Authority seems all for it, the response (page 1, page 2) from Larry McCallon, Chair of the Metrolink Board of Directors, pours scorn on electrification in general and on Caltrain's project in particular.  Some highlights:

Cost.  Chair McCallon: "Caltrain's 51-mile electrification modernization project is currently projected to be between $1.45 and $1.5 billion (infrastructure and equipment).  Metrolink operates on over 500 miles of track which would make this option very cost-prohibitive."

Zing!  He does have a strong point, in that Caltrain's electrification project is probably the world's most expensive electrification program, per route-mile.  Caltrain can evidently afford pre-construction cost blowouts that Metrolink can't.

Schedule.  Chair McCallon: "Caltrain's experience shows them to be behind schedule in their 24th year of planning the electrification of their 51 mile segment between San Francisco and San Jose."

Double Zing!  He is of course referring to old studies of Caltrain electrification dating all the way back to the early 1990s, and pointing out that Caltrain's planning process is just now coming to fruition.  In Caltrain's favor, this is largely due to a lack of money and political will, and not to any technical obstacles.

Shared Corridors.  Chair McCallon: "Caltrain, even with very limited freight service on the San Jose to San Francisco line is struggling with electrification compatibility with freight trains.  Metrolink, on the other hand, operates on shared corridors with much more frequent Amtrak passenger and freight trains that carry some double stack cars.  The electrification vs. freight issues would only be compounded in these rail corridors.  Between Los Angeles and Fullerton our trains also operate on BNSF Railway owned lines."

This point is spot on.  The freight railroads are adamantly opposed to electrifying any tracks where they operate, even if they are not the owners of such track.  The last time that Caltrain tried to kick off a CPUC rule-making process to cover 25 kV electrification, in 2007 under CPUC docket P0706028, the process was promptly shut down by the freight railroads.  The CHSRA's effort to clear HSR electrification ran into similar opposition, and survives only because it explicitly sidesteps the issue of electrification over tracks used by freight trains.  While the contracting process for Caltrain electrification is well underway, on the regulatory front, we have... crickets.

Electrification versus freight is going to be a messy fight, one in which Caltrain appears to have no friends, least of all Metrolink.

Chair McCallon's lack of vision should be taken with a grain of salt: Metrolink is a struggling organization with sagging ridership, dodgy finances and a governance structure that makes Caltrain look like a well-run corporation.  Nevertheless, the underlying issue of compatibility with high-speed rail is at least as important down south as it is here on the peninsula.  Let us hope that Caltrain's blended system will blaze a good path for Electrolink to follow.