Wednesday, April 8, 2009

Speeding Up Amtrak's Pacific Surfliner

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

A couple of weeks ago, the Ventura County Star published an article on California HSR that we mentioned, but only in the context of a negative response to it. What sort of fell through the cracks are these intriguing paragraphs in the original article and their possible implications for Amtrak Pacific Surfliner (APS):

Because the stimulus plan defines "high-speed" as trains capable of traveling at least 110 mph, more traditional rail systems in the Northeast and Midwest might be able to qualify for some of the funding — as might commuter systems in California.

"Some trains in Southern California corridors are pushing 110 mph," said Darren Kettle, executive director of the Ventura County Transportation Commission. "Some of my peers in Southern California are looking at that stimulus money to improve the Los Angeles-to-San Diego line and get that up to 110 mph."

Pacific Surfliner: Service Frequency, Speeds and Punctuality

The Pacific Surfliner route is primarily a semi-local train service between San Luis Obispo and San Diego via Santa Barbara, Los Angeles Union Station and Anaheim. The full route is 350 miles long, but the schedule shows that only a single daily train serves all of it. The LA - San Diego section is served by 11 daily trains (12 on Friday through Sunday), Santa Barbara - LA by 5, Santa Barbara - San Diego by 4 and San Luis Obispo - LA by 2. Many trains skip some of the smaller stations.

To get a sense of the speed profile along the route, I looked at the line haul times for a slow semi-local train (#774 southbound), which takes 8 hours 41 minutes from SLO to San Diego. Typical average speed between stations is 35-55mph. The busiest section, LA - SD, takes 2h40min to 2h50min, regardless of which train you take, at an average speed of 45-50mph. Average speed for SLO-Santa Barbara is comparable, the slowest section is Santa Barbara - LA at ~36mph average.

The upshot is that APS isn't really time-competitive against car travel in terms of speed, so its modal share of total trips along the South Coast is probably less than 1% right now. Its popularity relative to other Amtrak services is probably due to population density, the fairly long distances, the cost of driving if you're alone and, avoiding the stress of stop-go traffic in some sections. For business travelers, the most important factor may be punctuality, which for LA - Fullerton (BNSF section) has improved from 92% in the 2007 (Aug YTD) to 95% in 2008 (Dec YTD) and 97% in 2009 (Mar YTD).

Unfortunately, Amtrak's figure for the route as a whole was only about 85% in the last 12 months. According to the Glossary of Terms link off that page, Amtrak defines "on time" as less than 10 minutes late for routes up to 250 miles long and less than 30 minutes for routes of 550 miles or more. Interpolating linearly, the threshold for the full 350-mile Pacific Surfliner route is presumably 17 minutes. However, since many Surfliners actually implement part of the route, it's possible that the the threshold is computed separately for each train. If so, it would be closer to 10 minutes for most. That would mean 15% of all trains arrived more than 10-17 minutes late at their final dstination in the last 12 months. These details are relevant because rail operators in other countries use their own thresholds, so the numbers aren't directly comparable.

The primary causes listed are train interference, i.e. lack of capacity and - somewhat alarmingly - tracks and signals, i.e. infrastructure failures and maintenance impacts. Note the relatively high fraction of passenger-related delays, which would be reduced with improved on-time performance generally (reducing knock-on effects at transfer points) and level boarding platforms. These three leading causes reflect 84% of the total number of minutes of delay.

The most troublesome section is Moorpark to LA Union Station (not LAX airport), mostly because of Metrolink has to maintain the infrastructure and run a lot of trains on a shoestring budget. The San Clemente - San Diego section is a close second, because NCTD has to do the same. Only the top three segments contributing to each type of delay are listed, they don't add up to 100%.

Amtrak Ridership Statistics and Oil Prices

In FY 2008, Pacific Surfliner was the nation's second most popular Amtrak service with 2.89 million passengers (up 7%), after 3.3 million for Acela Express (up 6.5%). However, the revenue numbers paint a very different picture: $51 million for APS (up 6%) vs. $468 million for Acela Express (up 16%). APS serves a larger number of stations per mile than Acela Express, many passenger trips are probably shorter. Nevertheless, the numbers suggest that US consumers - at least those on the East Coast - are willing to pay a hefty premium for rapid over conventional rail service when short-hop flights become unattractive, as they did last summer. For 2009, Amtrak is offering discounted fares on Acela Express in a bid to sustain ridership. This is a reflection of reductions in short-hop air fares in response to falling jet fuel prices.

The Pacific Surfliner trains compete primarily against car travel rather than short-hop flights, but gasoline prices are nearly as exposed to oil price volatility as jet fuel due to the relatively low level of taxation (compared to Japan and Europe). Diesel is exposed as well, but trains make more efficient use of the energy at comparable seat capacity utilization rates.

Any high speed rail proposal in the US must therefore consider forecasts of future oil prices. In the coming months and even years, prices will remain depressed as the world economy deals with the aftermath of a massive burst asset bubble in US mortgages and derivative products. However, in the medium and long term, prices will rise again as China, India and other emerging economies achieve higher living standards. This will be exacerbated by peak oil considerations, i.e. the notion that the world will gradually run out of easily produced oil going forward.

Eligibility for Federal HSR Funding

If you take a really long view, as anyone contemplating upgrades to rail infrastructure ought to, it's a fairly safe bet IMHO that prices of oil-based fuels are going to rise faster than purchasing power in coming decades - exactly the opposite of the long-term historic trend. That's precisely why California has chosen to hedge its future by building an all-electric bullet train network that can run off a wide variety of primary energy sources, including the renewables favored by CHSRA.

Prop 1A reserved $950 million for HSR feeder services, with a generous slice of that reserved for the Amtrak California routes.

However, the federal concept of HSR is broader than the one promoted by CHSRA. Rapid rail, with top speeds of 110-125mph and some grade crossings retained, is also potentially eligible for federal funding. The South Coast is arguably an excellent candidate for such an upgrade, even though CHSRA has planned bullet train service between LA and San Diego for phase II. That's because funding for phase II extensions will depend on the commercial success of the starter line, which in turn depends in part on effective feeder services. Besides, no direct bullet train service is planned between Anaheim and San Diego so the two would not exactly compete against one another. In addition, plans do not include bullet train service to Santa Barbara.

Note that as e.g. Caltrain has shown with its "baby bullet" semi-express service in the SF peninsula, it isn't actually necessary to increase top speed to attract new ridership. Passengers care much more about line haul time and punctuality. Nevertheless, HR 110.2095 redefines HSR as follows:

"The term ‘high-speed rail’ means intercity passenger rail service that is reasonably expected to reach speeds of at least 110 miles per hour."

In addition, only applications for capital improvements to meet this new federal definition in one the 11 federally designated corridors are eligible for any part of the $1.5 billion that HR 110.2095 allocates for the purpose. Note that FRA's map is out of date with regard to the California system, Texas T-bone and other proposals. Chances are, Congress will fix that in this year's omnibus transportation bill now that there's federal money on the table. I wouldn't be at all surprised if e.g. LA - Las Vegas were added, hopefully as a spur off the California network.

As p136 the Joint Explanatory Statement Division A - part of the conference report - for HR 111.1 makes clear, the $8 billion in the stimulus bill for HSR use the same definition to determine eligibility.

Fortunately, this is vague enough to give USDOT a lot of leeway: neither bill specifies that 110mph or more needs to be sustained over a long distance. In practice, that means Pacific Surfliner would be eligible even if 110mph could only be reached in one short section, e.g. in Camp Pendleton. However, I suspect USDOT bureaucrats would then need to see a strong business case based on realistic forecasts of incremental ridership and fare box returns as a result of significantly improved line haul times and ideally, the provision of terrestrial WiFi on Board (cp recent trial on the Amtrak Capitol Corridor route). In addition, they would presumably want insight into the opportunity costs of sticking with the current service parameters: demand for more highway lane-miles, continued severe exposure to oil price volatility and, productivity loss due to time spent driving.

Potential Improvements

To get a sense of what might be achievable with appropriate investment, I assumed the primary strategy for improving line haul times would be express service between the primary population centers along the way, i.e. SLO, Santa Barbara, LA, Anaheim and San Diego. Let's call this Pacific Surf Express in analogy to Acela Express.

Next, I figured an express service might achieve an average speed of roughly 2/3 of the specified top speed. That's a very rough model, no more than a first order approximation. At the present top speed of 79mph, that would translate to a minor increase in average speed from ~40mph over the entire 350 miles to just ~53mph.

Nevertheless, even that would worthwhile: 19min gained from SLO to Santa Barbara, a whopping 53min gained between Santa Barbara and LA and a further 19min between LA and San Diego. Closer examination reveals that the largest single gain (~23min) would result from investment in the six-mile section between Glendale and LA Union Station. Average speed there is currently an abysmal 11.6mph, perhaps because of wait states associated with congestion in the throat of LAUS.

Run-through tracks for FRA-compliant equipment would benefit not just Amtrak Pacific Surfliner but other Amtrak and Metrolink services as well. These would be separate from those for the bullet trains. Considering the final EIR/EIS was completed over three years ago, the problem appears to have been a lack of funding. In a joint press release last May, Gov. Schwarzenegger announced that the state of California would invest $290 million into this and closely related rail projects as part of the Strategic Growth Plan. Of course, the state's finances are now in worse shape than ever, so it's unclear if the related appropriation of prop 1B (2006) bonds will happen in 2009.

All the more reason then to re-label those $290 million as a state contribution toward making Amtrak Pacific Surfliner an HSR service in the federal sense of the word. Other sections with low average speeds include:

  • Moorpark - Van Nuys (37mph)
  • Burbank Airport - Glendale (32mph)
  • Fullerton - Santa Ana (33mph)
  • San Clemente Pier - Oceanside (50mph, Camp Pendleton section)
  • Solana Beach - downtown San Diego (36mph)

In most of these cases, double tracking and/or signaling upgrades would be the primary approaches to improve line haul times. In Fullerton - Santa Ana, noise mitigation and bypasses at stations may be needed to increase existing speed limits. CHSRA also has a vested interest in that, because the last section to Anaheim is too narrow to accommodate dedicated bullet train tracks and, FRA currently permits mixed traffic only if there is sufficient and guaranteed time separation. South of Solana Beach, a short tunnel between Torrey Pines and University city would rectify and shorten the route.

North of Burbank, two approaches are conceivable for a Pacific Surf Express: a second tunnel bore between Chatsworth and Simi Valley or, switching the route to run through Santa Clarita and Santa Paula (CA-126 corridor).

View Larger Map

The ROW along CA-126 also belongs to UPRR but is probably little-used and in fact, abandoned between Fillmore and Santa Clarita. It contains some tight curves that would need to be rectified in order for trains to run through this mostly rural valley fast enough to overcompensate for the greater distance. Note that both the regular Pacific Surfliner and Metrolink would continue to serve the Simi Valley. Once the starter line for the bullet trains is operational, a Pacific Surf Express service via the CA-126 corridor could add a stop in Sylmar.

Ludicrous Speed!

All of the above applies for an express service with a top speed of 79mph, except in the Camp Pendleton section (and that only to become eligible for federal HSR funds). If the entire 350 miles were upgraded such that average speed jumped from ~40mph for semi-local to 73mph for express service, there would obviously be even greater time gains relative to the present situation: 58 minutes less between SLO and Santa Barbara, 86 minutes less between Santa Barbara and LA Union Station, 14 minute less minutes between LAUS and Anaheim and 47 minutes less between Anaheim and San Diego. That last section would then take 1h20m and LA - San Diego 1h45m. Improvements on this order of magnitude would be game changers.

For comparison, CHSRA is promising 20min for LA - Anaheim and 1h15m for LA - San Diego via Riverside. Considering that dirt probably won't be turned on phase II of the bullet train network before the 2023-2025 time frame, wouldn't it make a lot of sense to avoid the regulatory complication of mixed traffic in the Fullerton-Anaheim section? Is it wise for California to equate HSR with bullet trains at a time when the federal government is offering money for rapid rail projects as well?

In closing, this video shows how long all Surfliners might have to be to satisfy demand, if only they were substantially faster. The second locomotive is only needed because FRA compliance adds a lot of mass to passenger trains; enable mixed traffic via appropriate signaling, buy some lighter cars and one loco will be plenty for a Pacific Surf Express. Note that smoke and other emissions from diesel locomotives will be down sharply once EPA Tier 3/4 locomotive engines become available and operators switch to ULSD. As for those infernal bells and horns - mercifully brief in this case - quiet zones and grade separations, please!


Alex M. said...

Great post, I was looking into this extensively last week because we were talking about what to do in the time before the LA-SD HSR line is built. Looking through all of Caltrans' documentation, the current plan is to dual-track the entire route, with sidings for freight at certain locations. This will not only reduce delays caused by freight, but it will also allow MORE trains to operate on the line. The upside of this is the second proposal for the line which is to run express trains between LA and SD. Whether this means 1 stops in between or none, I don't know. The only reason they can't do that now is it means having to eliminate one local train that they run now which isn't possible given the constraints of the system. They are slowly adding double track sections and it will probably be at least another 5 years until it is all double tracked. They will most likely have to tunnel through Temecula Canyon which will take time. But the upside of double track+sidings=express trains=no stopping=sustained speed. I think improvements to the track may allow speeds as high as 110mph. I sure hope so, an express trip just over 2 hours is perfectly plausible.

Rob Dawg said...

Let me guess. You've never been to either the Ventura or Oxnard or Camarillo or Moorpark Segments. Between Camarillo and Moorpark there have to be a half dozen uncontrolled at grade crossings.

And before you say "close them" understand the easements are the opposite of what you are assuming. The railroad only has permission to cross the roads and not the reverse as the farmers were there before the railroad.

jim said...

The Surfliner has a lot more potential. higher speeds and increased frequency through the OC will run smack into NIMBY opposition though so get ready for that.

Alon Levy said...

The ratio of average to top speed depends on the terrain. When the terrain is very flat, as in the CV, it's close to 1: CAHSR gives the nonstop haul time from LA to Sacramento as such that the average speed is 300 km/h. Conversely, when there are a lot of turns and slow zones due to mountainous terrain or urban speed limits, the ratio is lower - e.g. LA-SF is supposed to average 260 km/h.

In general, here's my estimate of which factors affect average-to-top line haul speeds for the Surfliner:

1. Average speed: the higher it is, the longer it takes to accelerate to it, so the lower the average-to-top ratio is. At a top speed of 79 mph, a good train should be able to run at top speed almost the entire time, and even line haul speeds of 60 mph are achievable.

2. Acceleration: diesel engines and FRA compatibility reduce acceleration, reducing average speed. Even so, given straight track some lines, like the Water Level Route, achieve 60 mph on some sections. To gauge the true cost of diesel, we can compare the Pennsylvanian between Philly and Harrisburg with a Keystone train making the same stops. The electric Keystone takes 1:35 to do 104 miles with 4 stops, the diesel Pennsylvanian 1:44.

3. Interstation distance: clearly, the higher it is, the higher average speed is. However, this effect isn't very large at low speed and high acceleration. Even at low acceleration the difference is fairly small. We can compare the New York-Buffalo route, where the Lake Shore Limited makes 6 fewer stops than the Empire Service. It turns out the speed difference is small - skipping four stops from New York to Albany only reduces travel time by 3 minutes; the two other skipped stops, Amsterdam and Rome, are located in higher-speed sections, and save 7 and 9 minutes respectively. On the Surfliner, each stop skipped seems to save about 2-3 minutes. Going down from 8 stops between SD and LA to 3 will then save about 10-15 minutes.

4. Track conditions: slow zones kill average speed, no matter what the top speed is. So do tight curves. The Empire Corridor can achieve an average of 60 mph west of Schenectady because in many sections it's perfectly straight. The Surfliner has straight sections too, mainly between SD and Dana Point, but it has an obscene curve north of University City, and runs through urban sprawl for most of its way, which reduces speed. The best that can be done here is building a bypass track following I-5 in University City, which should reduce route length by close to 5 miles and increase speed.

In general, I'd say that just fixing the University City curve and eliminating 4-5 redundant stops should shave 15-20 minutes from the line haul speed, and increase average speed to 55 mph, on a par with New York-Albany and not far behind Philly-Harrisburg, whose fastest train averages 65 mph. Electrification can shave another 10 minutes, by which time SD-LA's line haul time will approach what is planned for HSR.

jim said...

The equipment for amtrak california is stretched to the limit as well and there is a need for capital to purchase more trainsets. The fleet is maxed out.

BruceMcF said...

Note that having a plan to upgrade to a substantial number of 110mph route segments and then applying for improvements that are part of that plan would mean there would be no need to "touch 110mph" in the first stage of the plan to be "reasonably" likely to achieve 110mph+.

Rather, what is required is that the plan is a reasonable way to get to 110mph+.

With both the Cascades and the Chicago - St. Louis route starting on an incremental upgrade approach and this year applying for the funding to actually achieve 110mph on part of their routes, that in combination with an existing 110mph blueprint makes it a lot easier to argue that an incremental upgrade of bottlenecks is "reasonably likely" to (eventually) get you to 110mph.

And that's not just playing rhetorical games ... getting sections that will continue to be bottlenecks up to their target for the completed Rapid Rail project yields dividends up front, AND by reducing the "footprint" of the bottleneck in the travel times, increases the expected ridership gains from the investment in the 110mph segment.

Obviously, don't need to upgrade to Talgos up front, but the target speed of the curvy bits should be planned with Talgos in mind.

Bear in mind that even if California is not a swing state, Missouri, Ohio, Virginia and North Carolina are, so get an approach in line with that cluster and it becomes harder to for the DoT to find fault with the approach.

arcady said...

The problem with the Pacific Surfliner is that the investment in infrastructure hasn't really been adequate to meet demand, although things are slowly improving anyway. One of the biggest bottlenecks on the south route, the single track between Orange and Santa Ana, was recently eliminated, and they're slowly working on adding more double track in San Diego County, as well as a third track on the BNSF Transcon. I think the biggest, most cost effective improvement though would be to come up with an integrated service for Amtrak, Metrolink, and Coaster, with the latter two providing local service, and Amtrak perhaps skipping more stops, which would instantly speed things up, possibly by as much as 15 minutes. In the long term, it would also be nice to replace the long, twisty route through the canyon north of San Jose, possibly by a tunnel with a direction connection to the UTC and future trolley extension.

On the north section of the route, by far the worst bottleneck is the single track between Van Nuys and Chatsworth, and the single platform at Van Nuys. Since the LACMTA owns that track, it would be up to them to deal with it. Maybe Measure R funds could be applied here. It's by far the biggest improvement to reliability, and would allow the removal of a lot of padding from the timetable, such as the 15 minutes that northbound train 799 spends sitting in a siding waiting for two southbound trains to go by. The Chatsworth-Simi Valley tunnel is actually not too much of a problem, assuming that other bottlenecks are eliminated.

Oh, and all of the route can benefit from curve straightening, for example at Burbank Junction and Orange, which both require slowing to 35 or 40 mph.

Rafael said...

@ rob dawg -

who said anything about closing any grade crossings?

@ jim -

I wonder if OC is still as full of Nimbys as used to be. I saw a report about the mortgage crisis in Santa Ana recently. Houses that sold for $600k just a few years ago have been foreclosed on, vandalized and adorned with graffiti.

In any case, the appropriate thing to do would be to first find out what causes the NIMBY reaction. Is it fear of falling property values? If so, what is the basis for that fear? Noise, grade crossing safety, air quality, something else?

Then you figure out how to mitigate the impacts that more and faster traffic would bring by applying money, technology and improved regulations. The deal should be that a negotiated and weighted set of objective environmental parameters should not become worse as a result of the upgrades. One of those parameters should be the number of trains per day that do stop at nearby stations even if express trains do not.

Baseline measurements of those parameters would be done by someone both sides consider impartial. After construction, frequency and speeds would be ramped up and the parameters re-measured, again by an impartial organization. Exploiting the capability of the new infrastructure would only be permitted as long as the scalar metric derived from the measurements does not exceed the baseline value. This type of arrangement gives both sides more confidence that they will not end up holding the bag.

To enable express service in a dual track section to narrow for quad tracking, you need to elevate one tracks and lower the other. Then add a siding and side platform under the elevated and above the lowered through track, i.e. only stack tracks at all of the stations.

If you do this such that there is an additional 12-16' of vertical clearance, one or more roads in-between will be grade separated via a through-pass. Ideally the stations should be sited such that a major thoroughfare runs on the middle of the platforms. The station hall would be split in two halves, each giving access to both platform tracks.

The elevated track and station should be designed with visual appeal and appropriate noise mitigation measures, preferably ones that don't increase the visual mass of the structure (e.g. glass sound walls instead of concrete). The underground platform will need proper ventilation and preferably, only cleaned-up diesel engines should be used.

The biggest issue is the small gradient for heavy freight trains, around 1%, as this has implications for minor cross streets to either side of the station. With tracks both rising and descending, it becomes impossible to keep all of them open and more difficult to grade separate them with over- or underpasses. Streets crossing where the tracks are almost at their full elevation difference can still get a through-pass, albeit with reduced vertical clearance.

For San Clemente, a different solution would have to be found as the existing track there runs past beachfront properties. Depending on the gradient profile, it might be possible to put the tracks in covered trenches underneath the slow + emergency southbound lanes on I-5 between the north end of San Juan Capistrano and the north end of Camp Pendleton. Impose clean diesel requirements and provide adequate forced ventilation plus some emergency access locations.

Rob Dawg said...

Rafael said...
@ rob dawg -
who said anything about closing any grade crossings?

Robert did when he suggested upgrading the speed in these sections. I'm not being curt or obstructionist when I point out that he is glossing over far more than a century of carefully negotiated accommodations.

arcady said...

I wonder if farm crossings can be handled the way they are in the UK: there's a gate, normally closed, and a phone to contact the dispatcher. If you want to cross, you call the dispatcher, he sets the signals to red and unlocks the gate. It could work nicely for very lightly used crossings. Otherwise, automatic barricade crossings should in theory be acceptable for speeds of up to 125 mph, and that's the highest that's really worthwhile on that corridor anyway. Oh and by the way, on the Surfliner South portion, the line speed is 90 mph for large parts of the stretch between Sorrento Valley and Santa Ana. It's one of the few places outside the Northeast Corridor where trains can go faster than 79, as it still has the ATS installation left over from the Santa Fe days.

Rafael said...

@ rob dawg -

(a) I'm the one who wrote the post and,

(b) the FRA permits grade crossings up to 110 mph if there are four-quadrant gates, even up to 125mph if the crossing is hardened. That appears to be code for installing something sturdy enough to stop a car or truck approaching the closed crossing at significant speed. One hardening technology - sorry, can't find the link right now - is called a vehicle arresting barrier.

Some implementations look like a super-sturdy tennis net that is either rotated or elevated out of the road when it is instructed to do so by the railroad signaling system. I'm not sure if or how these systems would detect the presence of a vehicle already above them as they deploy.

Inside the poles on either side of the road are coils of steel bands that must pass through a slalom of roller guides as a vehicle impacting the net pulls on them. Each section of the bands is bent severely and multiple times in alternating directions as it traverses the roller guides, dissipating the vehicle's kinetic energy. The systems were originally designed for protecting military bases, embassies etc.

To be effective, such devices must be installed at some distance from the rail tracks. That means moving the traffic lights back, for both cross and frontage roads. The other limitation is width, the videos I've seen all show either one or two lanes. One the up side, I've read they only cost half as much (~$500,000) as a set of four-quadrant gates.

Note that while FRA permits such barriers for speeds up to 125mph, the applications of the technology to railroad crossings is still new. That is why FRA hasn't spelled out exactly what "hardened" means yet.

Whether the CPUC would permit grade crossings to be retained at those speeds for cross roads with agricultural cross traffic is unclear. Full grade separation is probably not economical in these cases. During harvest season, train and barrier operations might need to be modified.

Alex M. said...

A couple questions regarding the LA-SD section for anyone who can answer them:

1) In the near-ish future (within the next 6 years), do you think Caltrans will be done upgrading the entire line to double tracks? Is the plan to double track this ENTIRE part of this route at all?

2) If the entire LA-SD line is double tracked w/ sidings and more trainsets are purchased, will Amtrak run express trains with at the most one stop between LA and SD?

3) The locomotives that Amtrak uses for the Pacific Surfliner route are supposed to have a top speed of 110mph. Will it ever be possible that PS trains will be able to hit that speed anywhere on the LA-SD line? With the number of coaches that they use, is this speed not possible?

4) Will Caltrans/Amtrak be implementing PTC on the LA-SD route? If so, how much time will that shave off the line haul time?

5) Besides double-tracking w/sidings, new track capable of faster speeds, elimination of worst curves, through tracks at LAUS, and PTC, is there anything else that will be done on this route to get the line haul time down to 2 hours for an express train?

mike said...

As arcady notes, the Surfliner already runs 90 mph in ATS (automatic train stop) territory. It's an old mechanical system installed by ATSF decades ago.

This is pretty common knowledge among anyone that is reasonably familiar with the Surfliners. It might help to enlist a couple of the more experienced posters on this board to review some of the posts before "going live" with them.

There's a total of 59 miles @ 90 mph. You can see a list of the exact sections of 90 mph track speed here:,548651

Robert Cruickshank said...

OC has plenty of NIMBYs. However, I believe the route is now double-tracked through the trickiest part - along Lincoln Ave in Santa Ana, between 17th Street and Santiago Creek. Next time I'm down there (I usually take Metrolink or the Surfliner from LA Union Station to Santa Ana) I'll try and take a closer look.

Most of the existing buildings alongside the tracks in the Orange/Santa Ana region are either warehouses, light industrial, or in a few cases, just north of the 22, apartment buildings.

South of Santa Ana there is ROW to add a third track siding in places, and in Irvine almost all crossings are now grade separated, save for Sand Canyon Ave. I'm less familiar with the track south of that point.

Rafael also mentioned "passenger-induced delays". This is indeed a problem on the Surfliners. Southern Californians generally don't know how to ride trains. This was an especially troublesome problem during the gas price spike of 2008, where lots of new riders were using the trains. Folks often didn't get ready for their stop in time, and the lack of at-level boarding platforms along the entire station length meant that trains had to line up perfectly for those who needed assisted boarding from the few existing platforms.

BruceMcF said...

"2. Acceleration: diesel engines and FRA compatibility reduce acceleration, reducing average speed. Even so, given straight track some lines, like the Water Level Route, achieve 60 mph on some sections. ...

4. Track conditions: slow zones kill average speed, no matter what the top speed is. So do tight curves.

The investment for track conditions is improved track ... the investment for tight curves is super-elevation to what the general traffic can allow, then additional elevation with a tilt-train.

With luck, the Federal government will get toward an 80:20 match for these kinds of investments, which will leverage state funding allotments by a substantial amount.

Even though diesels can only use passive tilt, its still well worthwhile because of the slower acceleration ... a higher speed limit through the curve has a bigger payoff for a diesel because of the longer time it takes to accelerate toward its top speed.

Rob Dawg said... "Let me guess. You've never been to either the Ventura or Oxnard or Camarillo or Moorpark Segments. Between Camarillo and Moorpark there have to be a half dozen uncontrolled at grade crossings."

This post is about edging the Superliner toward being a Rapid Rail line, and the standard answer for a Rapid Rail line is "upgrade the crossing". Save the 125mph gates for a stretch that combines lots of grade seperation and very few crossings, and nice straight runs ... quad gates otherwise, and speed sensitive so they do not close in time to give a 110mph train a through signal, when its a 30mph train trundling along the line.

People rarely object to upgraded gates, since it cuts down the number of Darwin's Volunteers, which widely considered to be a good thing.

sportbiker said...

This discussion is all well and good, but let's cut to the chase: how does this affect the peninsula? :-p

Andrew said...

Hometown Santa Barbara to LA in 90 minutes? BRING IT!!

Alon Levy said...

Bruce: between LA and San Diego, the only major curves that something can be done about are the one at Dana Point, which can be eased, and the one right north of University City, which should be bypassed. The other curves pass through dense urban areas, so that reducing noise becomes more important than making it possible in theory to maintain 90 mph.

BruceMcF said...

Alon Levy said... "between LA and San Diego, the only major curves that something can be done about are the one at Dana Point, which can be eased, and the one right north of University City, which should be bypassed."

If curves can be eased or bypassed, that is fine ... all rail services benefit from that, not just the Rapid Rail service. Otherwise, an ability to go through the curve at a higher rate of speed raises the average trip speed as much by accelerating the slow sections as by allowing full speed to be maintained in the fast sections.

"The other curves pass through dense urban areas, so that reducing noise becomes more important than making it possible in theory to maintain 90 mph."

Being able to go through a curve that is a 20mph curve for passenger comfort on a regular train, at 25mph or 30mph instead, gets the train through the slow section that much more quickly. And the more quickly the train gets out of the speed bottleneck sections, the better the return on the investment in the higher speed sections.

In some cases, the deceleration and acceleration itself is a source of noise, and being able to maintain a steadier rate of speed is a noise benefit. But, definitely, if noise walls allow faster passage through an urban area, that is another priority area for early investment.

Invest in raising the speed on low speed bottlenecks as early as practical, it increases the payoff on raising the speed on the high speed sections.

Now, just as in ongoing work on the Chicago / St. Louis line, if the speed limit is for track conditions, or if freight and passenger operations are getting tangled up, those need to be seen to first.

But tilt trains can help with the average speed even without raising the top speed. After all, that is the point of the Talgo's in the Pacific Northwest ... they have been the standard 79mph limit to this point, and are just now looking to apply to raise some sections to 110mph ... however, with some switchbacking in their alignments, they still benefit from the tilt-trains allowing them to go through turns at a higher rate of speed.

Brandon in San Diego said...

"Interpolating linearly"

oh my goodness!

mike said...

Bruce - There is nothing that prevents a diesel from having active tilt. See for example, the LRC, the ICE TD, or the Class 221 Super Voyagers.

jim said...

@ ragael- I think it was the san clemente area - and any of the areas where the train runs along the beach next to beach front property... those folks have put up a fuss before and generally don't like, want, need, or want to pay for, any public transportation... But since the trains are there, we could increase them little by little so they wouldn't notice. You'd think people who have so much would be too busy having fun to worry about the train going by huh.

- as for positive train control and everything else. the answer to all of it is, the railroad will eventually do everything but nothing will happen quickly or without additional money.

- as for boarding, i have seen that californians are most lackadaisical when it comes to such things. ( to some extent that is what is good about california - we are laid back at least we used ti be before all the east coast people tried to take over and ruin it) But yes , mostly they take their time about getting on and off the trains. There is a lot of pressure on the conductors to keep the trains on time to the minute. But again, one of the drawbacks in general with all of this growth and such, is that our pleasant casual lifestyle is being replaced by a more frantic one. Not really a plus. And what can you do with them anyway? close the doors on them? Push them? leave them behind?

Adirondacker said...

mostly they take their time about getting on and off the trains..... pleasant casual lifestyle....

There's a difference between pleasant and casual and taking up other people's time. It's rude to take more time than needed when you are getting on or off mass transit.

Rafael said...

@ BruceMcF -

the JR Hokkaido 281 (scroll down) is a state-of-the-art active tilt DMU running on legacy narrow gauge tracks between Sapporo and Hakodate. It is very lightweight and uses a computer with a track geometry database to anticipate curves and their superelevation. Its top speed is only 90mph or so, constrained by an old JR rule that requires an emergency braking distance of just 600m on the legacy network.

Nevertheless it manages very respectable line haul times - 3 hours for 200 miles - by keeping average speed as high as possible, even in winter conditions. Relative to the previous service, it reduced line haul time by 47 minutes and virtually killed off the short-hop air route.

For an explanation of balanced vs. unbalanced superelevation, see here. Enabling mixed traffic through appropriate signaling and other measures would be extremely helpful in increasing getting passenger rail speeds even without major track work.

@ Brandon -

anything wrong with linear as opposed to other types of interpolation as an educated guess? Amtrak just provides two data points.

@ jim, adirondacker -

close the doors on them?

Yes. On-time performance means passengers need to be at the platform on time. Admittedly, that is easier to enforce once trains run more often.

Push them?


leave them behind?

Yes. Ora inglese (~punctuality) is key. The reason people arrive late at the platform isn't a relaxed lifestyle, on the contrary: it's that they expect trains to be late and prefer not to get hot and sweaty standing around on an uncovered platform. If the trains run on time, passengers will arrive on time and accept that they may miss theirs if they're late.

Brandon in San Diego said...


"Interpolating linearly" seems very much like a technical term for something very simple.

jim said...

It only takes one or two people to hold up the show. My pet peeve is always the one guy who doesn't know the unwritten "stand to the right , walk to the left" rule on the escalator. And many people, don't under stand the " you have to let people off, before you can get on" rule of both etiquette and physics. This is beyond the grasp the general public. i have to tell that from what I see from a transit worker/daily subway rider's point of view... make no mistake, society is doomed. It just a matter of time.

Rafael said...

@ Jim -

in Japan, they paint lines on the platforms instructing passengers waiting to board where to queue up. Contrary to what you may think, Americans are willing to do just that, just look at any airport check-in or security scan location.

The key is making it obvious that queuing is what they're supposed to do, while explaining that the purpose is to help improve on-time performance.

arcady said...

And on the train, it's really up to the train crew to keep passengers informed when their stop is coming up, and herd them toward the exits when it's time. On the northern NEC, there are some stations with short and low-level platforms, so they only open a few doors on the train. The competent conductors will walk through the train, figure out who needs to get off, and tell them exactly where to go when the train is coming up to the stop. I've see dwell times as short as 10 seconds (one person got off).

Spokker said...

"The key is making it obvious that queuing is what they're supposed to do, while explaining that the purpose is to help improve on-time performance."

They would also have to inform passengers waiting on the platform how many cars the next train has since it could be different from time to time, especially on Metrolink.

Also, Metrolink and the Surfliner stop at different ends of the platform with some overlap at some stations. You would have to specify which queue is which, not just slap some lines on the ground.

This is all doable, of course. Just thought I would add more detail.

Brandon in San Diego said...

Spokker is absolutely correct.... it's obviosu then that a key characteristic is that the same equipment, or havingthe same standards.... are used on the line. Different equipment ='s different door opening locations

BruceMcF said...

@ Rafela: "the JR Hokkaido 281 (scroll down) is a state-of-the-art active tilt DMU running on legacy narrow gauge tracks between Sapporo and Hakodate. It is very lightweight ..."

And we have Talgos with concrete blocks in the trailer-driver because it "needs to be heavy".

Existing freight rail interests will use the bludgeon of FRA "safety" regulations to avoid "excessive interference" from passenger rail traffic until an accommodation is reached. When blocked by a regulatory body captured by the industry it is regulating, you have to go to the real source of the regulations to get things fixed.

The balanced versus unbalanced link is significant of what, exactly, in the Surfliner context? Precisely how many miles of the path that you are talking being balanced for 60mph minimum speeds, let alone higher?

jim said...

No, I'm pretty sure we're doomed.

Rafael said...

@ spokker, Brandon -

so you think the Japanese use the same equipment and the same train lengths all the time?

It's not essential that people queue up in absolutely the right place. It's important that they queue up at all and let passengers alight before they themselves board.

But I agree, having multiple transit agencies all do their thing without sufficient co-ordination of timetables, platform utilization etc. leads to greater delays than necessary.

That's where having a single, statewide passenger rail operator would come in handy but that's not going to happen as long as Joe Average perceives the state legislature in Sacramento as a black hole for his tax dollars. The 2/3 rule on the budget needs to go so there's some clear accountability.

Spokker said...

"so you think the Japanese use the same equipment and the same train lengths all the time?"

No, and this response is why I hesitated to post what I did. I was just adding more detail to the discussion of how passengers board trains.