Open thread time! Feel free to jump into the comments below.
- Driving like a grandma: electric service started on August 11th. Initially and until September 21st, EMUs operate on the current diesel timetable, for which they are grossly overpowered. Therefore, trips are sedate, and the full benefits of electrification are yet to be realized. We can't wait!!!
- Wasted dwell time: per FRA safety regulations, doors may only unlock and be opened when the train is at a complete stop. The drop step mechanism that deploys from below each door must obviously be fully deployed for the door to be unlocked and opened. Unfortunately, the software that controls this operation also appears to require that the train be at a complete stop for the drop step mechanism to operate. The result is eight to ten seconds wasted by the cycle time of the step mechanism (see video) which becomes a series contributor to dwell time. For an all-stops local, that waste is worth a full three minutes of run time. A typical Caltrain user travels for about 20 miles, let's say seven stops each way-- that's 140 seconds per day wasted. Multiply by 250 work days, and a full-time commuter will waste about ten hours waiting for those silly steps.This can be fixed in software. As seen in the timeline graphic, allowing the step to deploy and retract while the train is in motion would remove this wasted dwell time. The step mechanism can be cleared to move only when the train is moving slower than 5 mph, and since the step itself does not extend over the platform (it reaches only 63.5 inches from vehicle center line when fully extended) there can be no hazard to passengers. We spent a lot of money to save 25 minutes of SF-SJ run time, so please, let's not piss away 3 minutes because of overly conservative door sequencing software.
- Spares ratio: full electric service from September 21st onward requires at least 14 trains, not including spares. As of August 10th, twelve have been documented via photographs and video as having arrived in California (see tracking spreadsheet). One of them suffered a collision and was returned to the factory for repairs last March, leaving 11 known to be in California today. Two more (most likely 313/314 and 315/316) are due in early September. Peak service periods, when all 14 EMU sets are needed in service, can be protected with diesels if allocated exclusively to express services where they can sort of keep up, until further EMUs are delivered to increase the spares ratio above zero.
- Costly fender bender: Early this year, EMU 311/312 suffered a sideswipe collision at the CEMOF yard and was sent back to Salt Lake City for repairs. Two cars were damaged (see photo) and Stadler had to fly in welders from Switzerland to assess structural integrity and the cost of repairs. Aluminum isn't like steel, you can't just take hammers and a blowtorch to fix collision damage. The manufacturer's recommendation is said to have been full replacement, a multi-million dollar proposition.
- Wheel flat spots continue: detracting from the high quality of the new electric trains, there seems to be a continuing fleet-wide issue where all non-powered axles are prone to developing flat spots. The affected axles are located on the bike cars, and under the end cabs; once you hear this pattern, you can't un-hear it. As utilization ramps up, and especially when the next rainy season hits, let's hope this issue doesn't devolve into an epidemic of sidelined trains.
Feel free to comment on this or anything below.
Posts
Serious series contributor!
ReplyDeleteLuke warm take:
Mount the step on the platform, and have it in it's passenger operation position during all day, and only retract it when freight trains have to pass. Or possibly if it can't be "active" while an express train passes by, move it while there isn't any train in the station.
Re diesels as temporary spare for electric trains: Can't remember how many diesels used to be in use, but if no locos has been sold off or otherwise decommissioned a train could have multiple diesel locos. Don't know if the locos have the required equipment to control multiple locos from one cab. I assume they do as it would likely be the same equipment as would probably also be used to be able to drive the train from an end cab in a passenger car.
Btw what about the single (iirc?) AEM-7 electric loco that has been seen on pictures and/or videos from Caltrain? If Caltrain has access to it and if it's compatible with whatever standards the old Caltrain wagons has, it could form an additional spare electric train. According to the Wikipedia page about the Caltrain modernization program an AEM-7 has 5MW power while the EMUs have 5.7MW power, so unless wheel slippage would occur it would probably perform comparatively.
Btw it would probably be a good idea for Caltrain to study the feasibility to multiple their existing diesel locos with AEM-7 electric locos for the Gilroy services. Or possibly just use two drivers. (The driver in the "secondary" loco can just watch the brake pressure gauges and generally feel the acceleration/retardation forces and combine that with route knowledge to adjust the throttle. Back in the days before multiple control this was common practice in lots of places around the world). If it isn't wouldn't be feasible to control those locos from each other it would make sense to install passive wires for the system the diesel locos uses, in the electric locos, and form trains with a diesel loco, an electric loco and then all cars including the one that has a drivers cab. That way each end of the train would have a cab that drives the diesel loco, ensuring both that it's driveable at all by a single driver if the second driver is missing, that no extra driver is needed SJ-Gilroy, and that there is no situation where a driver forgets to start accelerate or whatnot. Sure, the trains would be longer, but worst case just drop a car?
That's a good idea! I wonder if they want to keep the AEM-7s as rescue trains and potentially use those trains for freight operations (if needed) on the line.
DeleteHowever - two questions:
1. Wouldn't it be operationally simpler and more straightforward with maintenance if they stuck with one mode (diesel) and ran them in shuttle mode from Gilroy - San Jose and back all day? There is a need for all day service IMO beyond just the commuter hours. Less locomotives = less maintenance, generally speaking.
2. Which option would be better out of these four: A - dual mode locos; B - a two loco consist (diesel + electric loco); C - a battery and electric loco that recharges at stations and under the catenary (~60 miles of range max is needed); or D - a diesel shuttle that can also later on service Hollister, Salinas, etc? Caveat: assuming in the long run that the Gilroy - SJ corridor will be electrified eventually.
I am not sure the AEM-7s have ever run. I certainly haven't seen any videos on Youtube. At the link below, someone claims they tried and failed to get one of them working. I doubt they'll ever run again.
Deletehttps://www.trainorders.com/discussion/read.php?4,5853472
From Caltrain officials, the AEM 7 are fried. They can't use them anymore and isn't worth the cost to repair
DeleteAnonymous - S:
DeleteGiven that they have all existing trains "left over" maybe they could run two diesel locos instead of one in the few Gilroy trains in order to give them better performance?
From an environment perspective only using diesel where there aren't any electrification would of course be the best choice. The caveat is that the diesels will have to drag the weight of an electric loco.
I agree that Gilroy should have some sort of all-day service.
Anonymous & Shero:
Sad :/ Thanks for the clarifications.
TL;DR: I can't comprehend why the AEM-7 and ALP-44 locos aren't being refurbished/modernized rather than just being scrapped (or almost-scrapped as for the ones Caltrain have).
I haven't really understod what differs between the US AEM-7 (and ALP-44) to their ancestor/sibling, the Swedish Rc loco? Unless there are some really major differences that affect the cost of maintenance and operations, it seems reasonable to maintain and repair those locos "forever" (or rather until they are involved in some serious accident). Sorry for bringing up this topic over and over in various places, but as a comparison the only Rc locos that has been scrapped has been the few ones that have been seriously fire damaged. Not 100% sure but I don't think any has been scrapped due to structural damage. The first one has been donated to the national railway museum not that many years ago. Afaik no scrapping except due to accidents is also true for the Rc (and Rb, which was a similar loco but without semiconductor motor drive) related locos sold to Norway and Austria, and I think this is also true for the license built locos built in Romania (for Romania and most of former Yugoslavia). An exception is that someone in Germany thought it would be a good idea to buy a Romanian loco (25kV 50Hz) and convert it to for usage in Germany (16kV 16,7Hz like what all Rc locos run on in Sweden) but it required too much effort and I think that that loco ended up being a parts loco in Sweden. I.E. the parts fit even between the license built locos and the "original" ones made in Sweden).
Also the Romanian loco Softronic Transmontana is based on the Rc locos. I bet that that company could easily supply parts for upgrading the electronics and whatnot on AEM-7 (and ALP-44) locos to modern standards, and also supply parts to replace all wear items, making them as good as new.
Btw the Swedish state owned cargo operator Green Cargo has also modernized some of their Rc locos, where the modernized ones are called Rd. IIRC they replaced the electronics and whatnot. They also modernized some of their GM-NOHAB diesel electric locos, previously class T44, now called Td, and those modernized locos can be multiple coupled even if you mix diesel and electric locos. (They can also be multiple coupled with all non-modernized locos within the same traction type, i.e. Td can be multiple coupled with all GM-NOHAB locos (and probably all other GM locos too?) and Rd can be multiple coupled with all Rc locos).
Caltrain only bought 2 AEM-7ACs: one (#938) was to be for testing the OCS, the other just for spare parts. According to Caltrain’s COO, they were never needed or used. He also denied the hearsay that either one was ever “tried & fried”.
DeleteOn the first electric run, conductors and staff were also commenting about the excessive dwell time from step deployment. So they are aware and hopefully motivated enough to push for a fix.
ReplyDeleteA few thoughts for Clem on here:
ReplyDeleteWhat's the cost/benefit on how much time savings this software update can save compared to if we just went straight to level boarding (as per the updated regulations from the CPUC)?
What's more cost-effective in terms of time savings here, level boardings vs curve straightening? I'd love to see some calculations on minutes saved compared to dollars spent. For example, fixing the CEMOF curve costs $1B, but saves 5 minutes every trip for a commuter = $27,3972.60 per minute saved (assuming 2x a day, 365 days a year) for that one person. Divide that by say, 2,000 riders, and that gives us a cost of $137 per rider for that year. Assuming the rider is paid $50 an hour (or their time is worth $50 an hour) and an average additional delay of 10 minutes (5 out, 5 back) due to curves or lack of level boardings, that gives us an annual "cost" of the rider's time of $3,050. When framing it this way, we can argue that a 10 minute daily savings will give back that much to the worker, and thus the $137 per rider expenditure for that year is worth it because it "saves" us that amount...
Just an idea but a nice way to kind of contextualize the dollar amounts here when thinking about people's time!
Level boarding imo is way better than curve straightening, at least for now. They can aggressively reduce the padding requirement. https://www.caltrain.com/media/33444 There is a study being done and will be completed by 2025. One of the Caltrain spokeperson said its mostly due to freight's loading gauge that they're worried about (they should just fight tooth and nail for the betterment of the public)
DeleteI would love a curve straightening but I don't expect any speed upgrades throughout the whole corridor even if they do so. The demand is not there for a 110mph express/ limited service just yet. (And probably requires a quad track somewhere too, so throw that into the mix)
> What's the cost/benefit on how much time savings this software update can save compared to if we just went straight to level boarding (as per the updated regulations from the CPUC)?
DeleteNormally the answer would be to prioritize the software change. But this is Caltrain, which means LTK Engineering Services would charge 10 gazillion dollars, plus the 30 years to deal with FRA/CPUC regulatory issues.
Re cost/benefit: the 16 (SIXTEEN!) years of posts here have been about pretty much nothing else. Browse the archives, and ... cry your eyes out.
Delete(Meanwhile he last 30 years of Caltrain have been about spend spend spend spend spend spend and deliver ... nothing, until late 2024, maybe.)
PS Be careful with "dollars per minute saved". That's the sort of nonsense that has delivered one express train per hour at "peak", and one train an hour, TOTAL, off-peak, for the last 20 years, all the while being trumpeted as a "success" and "bursting at the seams". (That would be Highway 101, not Caltrain.)
You want dollars per NEW rider, or dollars per kg of CO2, or something, anything else!
"Minutes on the train" isn't who chooses or chooses not to ride transit, just as "acres" aren't what vote (well, except in the US Senate.)
Shero:
DeleteThis comes back to my suggestion of having the steps at the platform rather on the train. But in the case of level boarding, just have a bridge that is say a feet wide along all of the platform edge, that folds upwards after the last train in the evening, and folds down before the first train in the morning allowing larger freight trains. (Also this would only be needed on one or two platforms at stations where all tracks have platforms, i.e. dual track sections and also quad track stations with platforms both for express/HSR trans and local trains). Don't know if there will be that many such stations? Up to including Santa Clara the freight trains can use the platforms used by Amtrak trains. How far north does freight trains actually run?
Anonymous: IMHO It's actually a great thing to use cost-benefit for the lives of the passengers as a measurement on what improvements are profitable.
But it can also be counted in pure operational costs.
Saving say 10 minutes for each return trip means that you can have one fewer train set (including fewer staff and whatnot) for trains that runs every 10 minutes. Run the trains every 5 minutes and you save two trains.
Not doing these kind of improvements, that don't result in any increased infrastructure maintenance costs but likely rather slightly reduced costs, is similar to peeing your pants to stop freezing. (Had this been on Facebook it would had been a matter of minutes before someone tags the tag group Oops, that's my kink :D )
From an operational cost perspective the main question is what it would cost to pay interest on a loan to finance the improvements. That might make things less fiscally attractive.
Moving CEMOF and say leasing out a large part of the freed up space would likely be a net positive even with relatively high interest rates. Leasing rather than selling, and contracts that can be terminated with a notice of say 10-20 years, starting from 30-40 years from signing the contract, would likely ensure a decent income from those properties while still being able to reclaim it for future transport needs.
Don't know what the land value is in the empty areas south of SJ, but I would assume that it would be relatively cheap to buy new land for a "CEMOF 2.0". Buy a bunch of extra land, add a station and sell the extraland when said station increases the land value...
MiaM
DeleteI personally believe its more prone to damage if the steps are at the platforms rather than on the train themselves. Not to mention, each station might have a different design depending on if the tracks are curved like San Bruno. Why not just have it on the train themselves and put a sensor at the end to stop the movement if the gap filler touches the platform? Two other issues come to mind is it requires precise stopping location (GoA2) to line up the bridge plate if its on the platform. Lastly, the freight trains dont just run at night, they also run at day. https://youtu.be/mbcHwhhPMEg?t=772
I don't know where they're moving or hauling stuffs but last I heard was somewhere near 22nd st to Redwood city. They could just use a smaller loading gauge for the locos and call it a day. Fight the issue at its core, and not having passenger train adjust to some clunky solutions.
https://www.reddit.com/r/transit/comments/vml9fm/z%C3%BCrich_sbahn_trains_have_level_boarding_from/ <- I just love this reddit picture so much because it shows that it is definitely possible with the KISSes that Caltrain has.
I definitely agree on the CEMOF remarks, but do they own the land? I know in 4th and King, Prologis owns the land.
@Shero:
DeleteI wasn't clear enough. I'm thinking about just one single step along the length of the platform, and having it fixed in place all day. (In practice it would probably have to be a few separate sections due to maintenance reasons and whatnot, but still).
But then I assume that the small gap you will have without a filler would be considered good enough. Maybe that isn't the case?
Add an extra second to the cycle time for the digital speedometer, which calculates a rolling (heh) average of distance traveled over some interval in the past. Presumably the number is rounded up to be conservative.
ReplyDeleteAnother time-saver would be to upgrade the track to 90 mph. Caltrain staff have said that the challenge with achieving this is primarily maintenance work needed. This is especially challenging with electrification because now they have to shut off the power any time they perform more than the lightest of maintenance, restricting work to the overnight hours.
ReplyDeleteThey should have done all the track upgrades first and then electrification
DeleteRight?
DeleteI think Caltrain should really marketing the:
* Higher average speed for EMU trains.
* Shorter commute times
* Comparing baby bullet times against off-peak driving times.
Transit will always be slower than driving, but that doesn't mean it can't be sped up. Why can't the speed limit for buses be higher than for cars?
Most people just whip out their Google Maps and if the transit time is 10-20 mins slower, then no amount of frequency is going to sway people from driving or taking Lyft. But if BART could operate at 90 or 100mph on the East Bay, how often would Google Maps show time where it's not just better for those who live within a mile of a station, but now those who live 2 miles within a station will change their commute.
I'm personally looking forward to other residents pointing fingers at Caltrain's new speeds and asking their agencies for the same speedups. Tri-Rail in southern Florida has slightly longer distances than Caltrain and it would greatly benefit from higher average speeds needed to compete with driving given how most stations are nowhere near downtown or any significant density.
Re speeds:
DeleteWhat are the impacts for train-road vehicle collisions at different speeds?
I.E. is the outcome almost the same at 90 MPH as at the current speed, or is it worse for the train? (Not to diminish what happens to the occupants in a road vehicle in a collision, but that is IMHO a road question rather than a train question). Like will the train need more repairs? Is there a speed where as a rule of thumb you kind of only need to pressure wash the train and continue using it after a collision?
A major problem with Google Maps and similar services showing travel times is that they show walking times between your actual origin and destination for transit usage while for car usage it assumes that both you teleport between your car and the origin/destination and the car teleports to/from a parking spot. Don't know how to lobby to get this changed, but the current "car" times should be displayed as "taxi" times (or rather as "tax evation taxi" as it afaik doesn't take into account lanes that are only allowed for transit vehicles and taxis). "Car" travel times should include walking to/from a parking spot that statistically is available, and also driving to/from said parking spot rather than in front of your actual origin/destination.
Given that Google collects all kinds of data from it's users and given that it usually knows what is a parking spot and what is something else, it should be able to get a grip on actual travel times.
Also re travel times: If the transit vehicles is faster than the roads for cars, including slow moving traffic due to congestion and whatnot, it's very much possible for transit to be faster. In the case of Caltrain you would have to have your origin and destination within reasonably short distances from Caltrain stations that runs some sort of express service, but still.
For the trains that stop at all stations the dwell times and acceleration would have a larger impact on speed than max speed.
Upgrading from 79-90 mph would not save much time. Given average Caltrain stop spacing and the performance of the new KISSes, a local should spend a minute or less at top speed, an express 2.5-3 min. The time saved traveling a few km at 79 vs 90 mph is minimal. It would save a local only 2-3 min over the whole distance from Tamien to SF, an express would save about 3.5 min SJ-SF. Trips not traveling the whole line would save less. By comparison, going to level boarding and shortening each station stop by just 15 sec would save locals almost 6 minutes and expresses almost 3. Its a far better deal to go to level boarding, which would cost a lot less than rebuilding almost 160km of track to 90mph while saving expresses about the same amount of time and locals almost double (plus being better for wheelchair users, people with strollers or luggage, etc.)
Delete@Shero, no, electrification is the far better deal in this regard and should have absolutely been done first. The better acceleration from electrification is saving expresses about 7 min and locals about 23 minutes, far more than a speedup to 90mph would save.
In fact, given average station spacing, Caltrain could theoretically run locals at 210kph/130mph and expresses at 320kph/200mph (the trains would accelerate to that speed and immediately begin to decelerate, never spending more than a second at top speed). Despite HSR level speed (and assuming curves can be ignored) the impact would be to save just 6 min for and 12 min for expresses versus going 79 mph - electrification and higher acceleration is as good or better than HSR for Caltrain distance and service patterns.
Note my rough analysis of time saved was assuming an average speed along the whole corridor. There are obviously curves that have a lower speed limit today.
Straightening the slowest curves to bring their speed limit up could have big time savings. While an 11mph increase from 79-90 mph doesn't save much, an 11mph increase from say 33-44mph drops travel time much more, both because its a 33% vs ~14% increase in the curve, and because of the time saved not decelerating all the way to 33mph before the curve or accelerating from 33mph back up to top speed after the curve.
The baby bullet in improvement at 90 mph goes from 59 to ~55 mins or nearly 7% shorter time? How's that not helpful? There needs to be a constant pipeline of projects that reduce trip time, improve frequency, improve reliability, etc...
DeleteThis PDF has a fun little table that lists relative maintenance costs between class 4 (current 79 mph), class 5 (gets us 90 mph), and class 6 (that coveted 110 mph).
https://onlinepubs.trb.org/onlinepubs/trnews/trnews255rpo.pdf
The track maintenance costs increase by 11%. EMUs are lighter than old diesels, so the cost might be lower.
Sure, level boarding would provide more time savings, but it's also inherently a single project rather than 24 small projects, each with its own neighborhood reach out, ADA considerations, elevator modifications, and more. Even if the total pricetag is the same, the relatively simplicity sways towards towards faster speeds.
There are other ways to mitigate speeds through cant deficiency waivers, upgrading switches, etc...
The metric to focus on is the effective trip time, which is a combination of dwell times, headways, and average speeds. Cutting the headways to 30 minutes saves an average of 15 minutes, and a further reduction to 20 minutes would save another 5 minutes on average. This can be done simply through more funding / faster turnarounds, but importantly no additional infrastructure. Then level boarding would save probably 15 seconds per stop, which is ~5 minutes for the entire trip.
DeleteRunning trains faster is probably least cost effective for Caltrain riders, but it's going to be needed for when CAHSR come.
Infrastructure maintenance cost depends on other vehicle related things in addition to just the weight / axle load. For example 100% low floor trams usually cause way more wear on the tracks than partial or fully high floor vehicles. Don't know how the new Caltrain EMUs rate in this perspective, but still.
DeleteInterestingly some sort of study has shown that waiting for a vehicle makes time feel about 50% slower than being on board a moving vehicle. I.E. 2 minutes of waiting feels like 3 minutes on route.
P.S. forgot to write that I 100% agree that it's a good idea to always have improvement projects in the pipeline.
DeleteStadler recommended replacement of the damaged cars, but Caltrain opted for repair due to new cars likely taking several years.
ReplyDeleteThere's youtube videos and posts on the Caltrain discord where the automatic announcements don't play on some routes
ReplyDeleteExample: [Manual Announcements] Caltrain Stadler KISS EMU in Limited Service 515 Mountain View to Hillsdale
Has anyone else experienced this? All the announcements played on the two EMU trips I got this week.
Re "Spares ratio: full electric service from September 21st onward requires at least 14 trains, not including spares ..."
ReplyDeleteI get 13 excluding spares, using your same (sadly unrealistic, inexcusably unrealistic, but here we are, in 2024, still finding 1980s level of reliability unrealistic hereabouts) 30 second station dwell times (as if!) and 10% schedule padding. Maybe you regard sub-30 minute turnbacks as unrealistic; I don't, fool that I am.
(So yeah, expect to see a lot of diesels subbed in to run "express" service for some time to come, especially as there's no reason to expect fleet maintenance to not be North American, ie shit, and so no reason to expect normal first world fleet availability. Sigh.)
That aside, this is all overthinking things. With 11 trains, and a far more generous 40 second dwell time and 15% schedule pad, you can provide better service than Caltrain has ever provided, plus meet every BART at Millbrae every hour of every day, which is something some people do like to go on about.
PS I did encounter some EMUs in service today, blowing their FUCKING APPALLING HORNS at the 16th Street crossing in San Francisco, and wow, the future is EXACTLY LIKE the shitty shitty past. I'm afraid to take a ride, because I know the interior noise level (insanely loud PA; endless "carrrrshawn therrrr derrrrs ... are abaahhhht ... to close"; pro tips from the TSA) is going to break my Swiss-expectations heart. At least on BART or Muni I know from just looking at the junk vehicles from the outside that I'm going to be abused by the PA when I ride, but the Caltrain KISSes are going to be total bait and switch.
Re PA quality:
DeleteDon't know about these in particular, but in general a problem is that in some cases the quality of the recordings themself for the prerecorded announcements are crap.
It should be recorded in a good voice recording studio, and the audio for each vehicle type should ideally be signal processed to compensate for imperfections in the frequency response and whatnot. This can really make a huge difference.
Anyone who aren't keen on using a voice recording studio is better off using a decent speech synthesis than just recording some random member of staff using the microphone in their smartphone headset or something similar.
Re PA audio level: Having each announcement start with the same sound and having a sound that in some form ramps up the audio level makes a loud PA way less pleasant. Like for example just start with the word "next" - the N sound will naturally ramp up the loudness. Kind of the full opposite to a gun shot.
Side track comparison:
In Germany Deutsche Bahn has two people who do all the PA announcement recordings, a guy for the S-Bahn trains and a woman for all other trains. This is great as it's a familiar voice wherever you are in Germany. Metros, trams, buses and also trains operated by other companies have other voices though, but they are at least the same within the same operator / transit agency (i.e. the same voice for IIRC all metros, trams and buses that BVG runs in Berlin and so on).
In USA there isn't that much difference in accent between areas - I.E. anyone understanding the accent in one are will most likely have no trouble understanding the accent in every area, so it might not matter that much if local people do the recordings in each area. But for a country like Germany or for that sake the UK it can be really hard at least for visitors to understand every accent, and thus a common "railway voice" is a great idea.
Taking another side track:
DeleteI like how SF Muni has a female voice for eastbound announcements and male for westbound trains. This would be fun for Caltrain onboard announcements to distinguish northbound vs southbound trains.
I'm finally on one now and I mostly disagree about the PA. The volume and audio quality are much, much more pleasant than the loud and scratchy manual announcements on the old trains. Yes the "caution doors are about to close" is mildly annoying. But, like, *mildly*. On net I'd say it's a big improvement.
DeleteWifi isn't working unfortunately. Seats are comfier than I thought they'd be from the photos, but I wish the seatbacks had a couple more degrees of lean. Info screens are very nice. Upper deck is much nicer than the European KISSes with their shit loading gauge. Aircon actually works. For all the missteps, I'm happy to be out of the 1950s.
This was in reply to a comment from Richard which seems to have been deleted for some reason
DeleteAnnoyingly this platform lets comments be published and replied to, and later decides that a comment is spam so I have to go fish it out of the filter.
DeleteThe Caltrain/BART Schedule Coordination (PDF, 1.58 MB) presentation for this tomorrow’s Caltrain JPB CAC meeting contains:
ReplyDelete1. Electrification Schedule Overview (with string chart)
2. BART Service Changes
3. Millbrae BART Transfer Optimization
4. Selected Other Transfer Timetables (VTA, ACE, Capitol Corridor)
Well, I'm sad that Caltrain couldn't get local service down to 75 mins. Is it the extra dwell time from the slow step extension + door opening that moved our expected 75 mins to 77 mins?
DeleteVTA transfers are absolute crap. If VTA shifted their arrival time to be 2 mins earlier, they could have many more shorter transfers. What's worse, I was once on a nearly empty VTA train that sat near Moffet for 3 mins because it was running ahead. It ended up arriving at MV just as train was departing.
In the evening in MV, if VTA shifted their departures to be later by 2 mins, sons of connections could be made in 15 mins fewer.
Caltrain to BART in the morning is pretty bad with only 2 out of 4 Caltrain arrivals having a good connection. Caltrain is being shady here by hiding 1 out of the 4 trains. Unclear why you need a whole 5 mins for a "GOOD" transfer despite the southbound transfer requiring a 2 minute run up and down stairs.
Weekend locals at 79 minutes SF-SJ with 21 intermediate stops. Extraordinary - they've almost matched the 1978 SP timetable which managed 75 minutes SF-SJ, also with 21 intermediate stops!
Deletehttps://wx4.org/to/foam/maps/2Wx4/SP_003/1978-04-01SP_PeninsulaPTT.pdf
Of course you can come up with excuses/reasons for this (higher ridership; how knows how on-time those SP trains were; etc.). But it's still interesting to put into perspective. I will also note that the 1978 timetable has the fastest all-stops local SF-SJ of all the timetables I checked dating back to 1866. So 79 minutes is still somewhere around second place.
It's news to me that the expresses will be peak only, "generally covering 6:30AM-9:30AM and 3PM-7PM." That's unfortunate.
DeleteIn the old, waived, 6 trains-per-hour commitment made to the FTA for electrification/rolling stock funding, do we know if that was peak only or all day?
6 tph was only ever a peak time requirement (core capacity). It's fun to look back at how Caltrain used to be bursting at the seams, and opponents were using their interim capacity increases to undermine the FTA funding for electrification! My how times have changed.
DeleteI'm not sure I trust that 1978 SP timetable. It shows 13 min SF-So.SF regardless if the train makes 0, 1, 3 or 4 stops. Something doesn't add up. The "75 minute 21 stop locals" are all trains that cover SF-SSF in 13 min despite making 3-4 more stops in between than the expresses that do in 13 min as well.
DeleteAbout the 1978 timetable: I wonder if some of those stops were flag stops that got bypassed if no one was getting on or off
DeleteOh gosh.
ReplyDeleteThe transfer times are horrible. Like if they intend to run 4TPH rush hour (which in itself is a horribly low goal - that can be achieved on a mostly single track railway...) they should be able to have less than 40 minute wait time to change between Caltrain and ACE.
Even needing to discuss wait time when transfering between a fully at least double tracked commuter railway with a light rail line is embarrassing.
Maybe it's just me, but does anyone else feel the String Chart lines need to indicate the signal system limitations? Otherwise, you're showing trains getting closer, but we don't see whether we're approaching any limits to how close they can get.
ReplyDeleteGood idea. Wouldn't be that hard to add signal/block sections on the charts.
DeleteSomeone pointed out elsewhere that the deployed steps violate AAR Plate F and this precludes their deployment when not docked at a platform. However, a train that accelerates or decelerates at 0.5 m/s^2 (one twentieth of a G) will only travel five meters between stationary and 5 mph (sorry, mixed units) which is less than the distance between the cab and the first door. Effectively, any train moving at only 5 mph has all its doors at a platform so this concern is unfounded.
ReplyDeleteThat simple arithmetic proof is beyond the grasp of engineers that are taught (usually for an OK reason) to slavishly adhere to the most conservative guidelines. Works great if the problem is to determine the strength needed for a beam to support something, where nuance is less necessary. It's much easier for them to say train must be stopped to make certain no violation of AAR Plate F occurs. Why look for the nuance? Rules must be respected.Let's not go picking nits or examining nuances to just save a few seconds. ...and don't even get me started on fire "safety" issues...
DeleteThis issue of waiting with doors closed for the moving step to deploy/stow while stationary adding to station dwell times (and therefore end-to-end run times) was raised with Caltrain’s COO yesterday.
DeleteUnsurprisingly, and missing the point, the response was we’ve built that time into our schedule … followed by the FRA probably wouldn’t allow moving the steps while in motion, and that they might “clip” someone that’s “too close.”
Since the steps appear to clear the platform in Clem’s linked video clip, it’s unclear how they could possibly, realistically “clip” anyone on it. Also, while CA HSR’s current train RFP specifies trains nearly 15 inches wider than Caltrain’s KISSes. (3,378.4 mm vs 3,000 mm), Caltrain’s COO said he was told by HSRA officials that the trains being acquired as part of the current RFP for the IOS “will never make it to the Peninsula.” Huh!?
"Caltrain’s COO said he was told by HSRA officials that the trains being acquired as part of the current RFP for the IOS “will never make it to the Peninsula.” Huh!?"
DeleteThat's been obvious for decades.
It's equally obvious that the California High Speed Rail contractor-enrichment scam won't need any sort of high speed train for decades -- not until either SF-Bako or LA-Bako is completed, which is not for decades, and most likely never. I mean, look at their "progress" over the last three decades! Just look!
Any "high speed" trains — extra special, extra US-consultant-enriching, triple-global-cost, snowflake-enabled — "high speed" trains they procure any time in the next decades are going to be end of life well before the tracks they could use are constructed.
Pretty cool deal if you're a US engineering consultant mafioso, though! Pretty cool!
Assuming whatever bullshit concete they're erecting in the Central Valley ever gets any superficial layer of ballast and tracks on top of it, the trains that any realistic or rational person might run on it look like this (assuming electrification of the useless stranded, deliberate worthless, maximally contractor-enriching Central Valley concrete wank-fest, a mindless exercise that would just be adding more worthlessness on top of uselessness, so of course they'll electrify an isolated nowhere zero-ridership line, for great profit) or this -- not remotely like this.
I doubt any of are going to live to see high speed trains on the peninsula, except maybe towed in and then run back and forth a couple times as a cash-seeking stunt. I certainly won't. (Hey, I'm old enough to have seen an ICE-1 diesel-hauled up and down the Caltrain line in 1993. History doesn't just rhyme, it actually repeats.)
It's sure super profitable to delude others into imagining otherwise, but it has nothing to do with California in 2024 -- over thirty years into the "state" CHSRA's consultant-enrichment adventure.
So, no, no CHSRA IOS RFP trains for you, SF Peninsula. Not ever. Nor for you, LA Basin.
"This issue of waiting ...adding to station dwell times (and therefore end-to-end run times) was raised with Caltrain’s COO yesterday. .. Unsurprisingly, and missing the point, the response was we’ve built that time into our schedule …"
DeleteCool beans! It's great to have a the inside scoop.
This technical explanation from America's Finest also builds endless breakdowns into the schedule!
This technical explanation from America's Finest also builds not implementing level boarding during the last 30 years, and absolutely not planning to do so for the next 30 into the schedule!
This technical explanation from America's Finest also builds conductor-operated doors into the schedule!
This technical explanation from America's Finest also builds third world track maintenace practices into the schedule!
This technical explanation from America's Finest also builds we're the ones who brought you CBOSS, or whatever 20x 0.1x functionality rent-seeking private-profiteering globally-useless signal shit we ended up reaming you with into the schedule!
This technical explanation from America's Finest also builds endlessly protracted terminal station turnback dwells into the schedule!
This technical explanation from America's Finest also allows builds insanely low terminal approach speeds and crap trackwork into the schedule!
This technical explanation from America's Finest also allows builds "can't run any more trains" because we triple the number of staff required to run each train into the schedule!
I know I can't say "death is too kind a fate".
But I can't help but speculate whether a human exhibiting this low level of brain stem function is really capable of experiencing any quality of life.
I do honestly try to close my bold and italic tags. I honestly do try. I try.
DeleteThe solution here, as I have stated before, is to require that both Caltrain and CAHSR (and Metrolink, Capitol Corridor, Brightline (East and West), etc., etc.) use trains that are 3,251.2mm wide per plate F. Then with your platform 1701.8mm from centerline you have level boarding for all services without requiring any retractable steps, their cycle time, or their maintenance.
DeleteClem, but at 0.5 m/s^2 a train will reach 5 mph in just 4.5 sec, when you identified the cycle time of the steps as 8-10 sec. Plus, Caltrain's new equipment accelerates at 1 m/s^2 (correct?) which means it would cover 32m-50m while the stairs go in and out.
Ignoring my actual solution above, it would seem the hack would be to define a "Plate F+" clearance for 100m before and after the platform where the train can be wider than the loading gauge on the rest of the line while steps are going in and out during acceleration/deceleration as you suggest.
Unfortunately passengers getting hurt isn't unheard of.
DeleteIIRC a decade or so ago someone got a leg clipped related to the train running Stockholm-Uppsala, which at the time was running using 1980's loco hauled cars, with doors at the ends of each car, obviously with the between-cars gap next to the door, forming a space where someone might fall in to if they slip.
Nowadays I'm not sure which trains are used for this. Some of the traffic has moved from SJ (the equivalent to Amtrak, sort of, but only running trains but not owning any infrastructure (except perhaps some shed or maintenance facility or so) to the various regional transit authorities. If it's still SJ I hope that they rather are using the class X40 "berlin wall" two level trains (called so due to them being gray and tall :D ).
Anyway, IMHO the solution to both slow doors and slow steps is to use pneumatic operation and also light weight materials. That way both the step and the doors can be slammed out/in opened/closed without any risk of major injury if a door or the step hits someone while in motion. Of course for fast trains there is a need for some additional mechanism that locks the doors shut when closed, otherwise for example a train running in the opposite direction on and adjacent track might cause enough drag that the pneumatic actuator might not be able to hold the door shut. (I've seen doors open a few inches due to such drag, iirc it was on the 1980's class X10...X14 EMUs).
It sometimes seems like certain simple technologies are lost in peoples minds.
If electric operation of doors and steps are really really a requirement then just use some sort of sensor that detects if there is more resistance than expected, and if so move slowly or alert the driver (who could just monitor it via CCTV and press a go-ahead button if it's just drag due to lack of maintenance, dust, dirt or so rather than a human getting crushed).
Also I think that we need to define a level of acceptable injury due to minor accidents. I.E. for example having to go to a doctor to have your nail punctured to release blood building up underneath it due to you being stupid and trying to force closing doors to not close seems like an acceptable level of injury. It will eventually heal and a few months later no-one could tell which finger got crushed.
But also: Having more frequent trains stops problems with passengers getting injured while trying to board a train that is already closing it's doors, or even trying to hang on to the outside of the train if the doors are already closed. (Something that would also solve this, but is way harder to implement, would be a society that aren't so focused on being perfectly on time to everything).
This comment has been removed by the author.
DeleteWhat about deploying/retracing the steps at the same time as door opening/closing?
DeleteThe SF Muni older Breda LRVs deploy their gap filler in a flash... Sadly, the new Siemens are glacial and between slower door operation, they easily take 5 secs longer to open and close them at the stations.
DeleteSlow door operation seems to be a new industry trend...
Well, OK then!
ReplyDeleteCaltrain, the most successfullest Public Transportation Agency Ever in the History of the Universe, staffed and abetted by America's Finest Transportation Planning Professionals, has, here the the year of our lord 2024, revealed unto us the VERY FIRST NON-TRIVIAL CALTRAIN TIMETABLE UPDATE SINCE 2004 (no joke! since 2004!) and ... Clem's high-school-level-mathematics (no insult, none at all, just the facts!) "Taktulator" numerical integration runtime data from 2011 or before precisely model the service Caltrain proposes to operate in late 2024. (No, no late 1998. Note late 2010. Late 2024. 2024! Adjust your chronographs!)
Caltrain's published proposed all-kool all-electric late-2024 timetable, after more than three decades of delay and tens of billions of incinerated public cash, literally incinerated!, corresponds most excellently with THIS Taktulator scenario (10%!!! schedule padding, uniform 25 seconds (!!!!!!!) dwell time)
Science. It works, bitches!
(Tens of billions of dollars of fraud and graft and rent-seeking and plantary incineration? Yeah, that works, too. That works for some people. In a very very very different way.)
PS 10% schedule padding? 25 second dwell time? With no level boarding? Sure, Jan!
Ignore/delete this. Superior model (15% fat, 25/30/45s dwells) posted just below
DeleteCaltrain published their new timetable, which marks the very first change in service (other than steadily slowing service and off-peak cut-backs) since 2004.
ReplyDeleteTaktulator (code and data from 2011!!!) nails it at 15% schedule padding, 25 second dwells for all-stops local (2tph) and 45 second dwells for limited stops (1tph or one and pointlessly 1tph of another, because some things DO have to stay the same as in 2004.)
Taktulator: 2024-09-21 25/45s 15% 6MW
Fleet size: 14 trains (not including spares)
Fleet utilization: 67.7% (gotta keep doing things the Commuter Railroading Way; all Commuter Railroader's trains, no matter how shiny or expensive, like to take long long long naps all day long)
34 minute turnback times in SF, 25 or 47(!!!) in SJ, 35 at Tamien.
Symmetrical northbound and southbound (excluding the Gilroy boat anchor, which chews up four diesel trains and presumably four trains crews all by itself), for the first time in Commuter Railroading History.
1tph to Tamien, man what an incineration of cash wiring that. San Jose gotta San Jose.
Caltrain 2024-09-21 weekday PDF
DeleteCaltrain 2024-09-21 weekend PDF
2024-09-21 weekend 30s 15% 6MW (Fleet 8 trains; Score 104)
DeleteOf note (good note!!): uniform 30 minute headway, and (apparently, unless they left off some annotation) identical Saturday and Sunday service, for the first time in Commuter Railroading History.
HOORAY!
Of other note: run time a couple minutes slower than the weekday locals, with station timings annoyingly (but Very Commuter Railroading-ly) slightly different on weekends than weekdays. Taktulator nails this as same 15% schedule padding as weekdays, but with dwells modelled as 30 seconds versus weekday 25s.
I imagine the Commuter Railroading Thinking is that nobody's going to ride the 2tph all-stops trains during weekday peaks because people are going to time their existences around a one limited-stops "bullet" straight outta 2004, with modelled dwell times twice as long. My expectation is that this will be less the case than in the bad-schedules past, but we'll see.
And it's not as if "connecting transit" is even a thing here in the worst of all possible worlds, so ... no matter. The only effect it to make printed timetables larger, and to make a few people miss their train and have to wait 29 minutes. I'd have identically sandbagged the weekday local dwells, but that's just weird Galactic Takt Brain talk.
Turnback times (40ish minutes) and fleet/crew utilization (67% ugh ugh ugh) are dire.
arrive SF :15 depart :55 / arr SF :45 dep :25
40 idle minutes, twice an hour.
One could save a train and crew if this turnback, modelled as 15 minutes minimum, and mostly likely Commuter Railroaded at Caltrain to be required to be 30+ minutes, were squeezed to 10 minutes, which, believe it or not, Caltrain has scheduled a few times over the years only at SF only at weekday peak only. I don't see them managing this today, or, sadly, any time in the future. Caltrain service is and will remain just too unreliable, even with unprecedently-low 15% padding, for all the usual sad reasons.
arr SJ :43 dep :26. 43 minutes non-revenue, 1tph.
arr Tamien :18 dep :51. 33 minutes non-revenue, 1tph. Brief nap, by their standards.
Two thoughts:
Delete1) deleting the express and replacing it with another limited increases the timetable score by a whole two points, largely because it reduces maximum wait times throughout Silicon Valley. The transformative effect of regular 15-minute takt on connecting busses and shuttles would be awesome, but the continuing obsession with sub-hour SF-SJ runs (even as the average Caltrain trip is ~23 miles) sure gets in the way.
2) I regret not calculating a 6 MW 79 mph performance table, which back then Richard and I thought would be silly, since why use so much power unless you can go faster... I've tried to resurrect the old code that I wrote and got bogged down trying to exactly reproduce the old results. I guess it's not crazy to swap in the 100 mph performance table because there are so many stops it'll almost never reach that speed.
Oops, your comment re 6MW-100mph made me double-check my work and I see I'd accidentally typod the "Santa Clara Limited"/"Silicon Valley Express" ("SVX"! Branding!) 1tph limited as "KISS-100MPH-4MW" when I made everything else 100MPH-6W. I've updated that, and updated its modelled dwell times to a schizo combo of 45s (same as "Baby Bullet" 1tph throwback) north of Redwood City, 25s (same as 2tph weekday not weekend local) south of RWC, and popped in 45s of extra dwell at RWC to minute-level sync up with the local takt.
DeleteUpdated: 2024-09-21 25/45s 15% 6MW (Fleet 14 trains; Score 117)
Re "not calculating a 6 MW 79 mph performance table": it is indeed kind of nuts, and we'll see how long they can keep it up. The energy costs are going to bite, and I imagine train lengths are going to be "right sized" (ie intermediate cars spliced out of 175m super-sized trains, not to make new trains, which they can't do because of "client from hell" Caltrain's insane bespoke very-not-Swiss train power configuration, and plus there are zero spare cab cars, but just to sit growing moss, but not sucking power) at some point to try to stem the bleeding a tiny bit. Go fast and draw juice to try to make up for uncontrolled dwell times due to ... no level boarding. It's all so so so stupid.
BTW there are few other glitches (obvious bugs where a faster and or more powerful train has slightly longer trip time than a lesser) in the performance tables that I've spotted over the years, so it's not as if your over-simple and but excellent 13+ year ago Octave(?) numerical integration was perfect. It's pretty amazingly good through! Still, it would be nice to add 6MW-79MPH, even leaving the old tables untouched for bug-for-bug backwards compatability, in consideration of the hundreds of thousands of avid taktulator users galaxy-wide and their expensive software investment.
Don't remember if the new Caltrain trians have regenerative braking or not. If they do then the energy cost might not be that high. (With AC electrification at mains frequency the regenerative braking feeds back into the grid). Sure, there will be losses and whatnot, but still.
DeleteBtw do railways pay the same price for electricity as other consumers?
As a comparison, IIRC there are no taxes on electricity for rail in Sweden. (Can't remember the details but diesel either has lower or no tax at all).
Re "deleting the express and replacing it with another limited increases the timetable score by a whole two points, largely because it reduces maximum wait times throughout Silicon Valley. The transformative effect of regular 15-minute takt on connecting busses and shuttles would be awesome":
DeleteAll true! Also if one shifts either the entire northbound or entire southbound schedule a couple minutes, the glimmers of hints of takt "symmetry points" vaguely start to appear at all of Redwood City, Palo Alto, San Antonio (irrelevant), Sunnyvale and Santa Clara. That's due of course to a lucky rough coincidence of 7.5-ish minutes stop-to-stop between those stations. Imagine what could be done if anybody with a clue were involved in any way in deliberately engineering such "coincidences"! (Oh, hang on, that's the last 16 years of this blog ...)
Taktverkehr is just the kindergarten level.
"Integraler Taktfahrplan" is the REAL deal.
"Takt und Symmetrie" is the real slogan.
https://de.wikipedia.org/wiki/Symmetrieminute
(The wonderful Marco Chitti, who for unknowable reasons still posts only on the Nazi hellsite, once wrote a really nice simple kindergarten-level intro in just 20 tweets.)
Re "the continuing obsession with sub-hour SF-SJ runs (even as the average Caltrain trip is ~23 miles) sure gets in the way."
DeleteYeah, that "baby bullet" sure is a harmful throwback.
Aside from(!) all the other continent-wide Olde Tyme Commuter Railroading toxic baggage that screws us, a further deadening factor with Caltrain is that they never made any substantial change to their timetable in the twenty years between 2004 and 2024, and so the place is stuffed with lifers who have literally known nothing else, some of whom notably have never even really worked anywhere else!
Mix that with endless smug self-congratulatory ridership reports ("stonks go up and to the right! Up and to the right!") which in reality measure nothing but commute-hour 101/280 traffic congestion, and you have a recipe for more of the same, forever. Reporting to an easily manipulated governing board which has always had a working majority of utterly disinterested, downright corrupt, and/or plain stupid individuals guarantees under-achievement will only be rewarded. Can't mess with "success"!
But hey, 2024!
Something improved, a bit!
If they do manage to operate this for-once-okay-adjacent timetable with any level of predictability then it's hard to see the counter-productive 1tph "baby bullet" hanging in there against riders who get a glimpse of a non-abusive timetable and ask for even less 1tph punishment.
It's all ready and waiting ....
According to prior staff presentations to the board, power costs have skyrocketed since electrification was planned. So instead of saving money vs. buying diesel, staff was projecting nearly a doubling of propulsion energy expense to run the new schedule.
DeleteLast public update was that they were working well with Peninsula Clean Energy (SSF OCS substation) and SJ Clean Energy (SJ OCS substation) on obtaining a slightly more favorable “transit agency tariff” on their green (carbon-free) electricity.
Caltrain’s COO was asked a couple months ago about the possibility of “right-sizing” (shortening) some 7-car trainsets for more lightly-patronized off-peak runs (as BART has done to save on energy and per-car-mile costs and to lessen rider isolation for improved on-board safety).
The short answer was “we’ve looked at that and with the way the trains are configured it’s just not possible.”
Also, staff reports that the WCs automatically flush a total of 3 (yes, theee!) times per use and that when the retention tanks reach 80% full, the WC takes itself out of service. This has already started happening (to the chagrin of riders needing & expecting the ONE WC per 7-car train to be available) because the retention tanks are “only so large.” So staff says they’ll probably need a “honey truck” at both terminals to pump the retention tanks to avoid surprise WC shutdowns. Nice, eh?
They acknowledge one WC isn’t enough, having already thought about installing more someday, but will possibly wait some year(s?) to do so because it would invalidate Stadler’s warranty on each entire trainset so retrofitted. Meanwhile, staff is mulling whether and where to install Throne Labs or JC Decaux style toilets at selected stations.
Lastly, the “doors closing” warning announcements are annoyingly firing more than once throughout the train because conductors push “close doors” buttons at least twice: the two buttons for doors other than at the control panel door, and finally and lastly the one for the door that they’re using to check that all the other doors are clear & closed. Ugh!
It's not just electrified Caltrain's electricity costs. A very tiny number of us have been wondering for a long, long time about the "coming" future high-speed trains, and their costs to operate, given that they not only are energy hogs at speed, no matter how some try and fail to rationalize that away, but the route also features several intermediate stops between San Francisco and Los Angeles, one of the fundamental defects of the project since it first was undertaken decades earlier. Each of those stops costs time, but also energy for re-acceleration to normal operating speed. Such penalties were the subject of at least one paper in more recent times. As it is, I have wondered if in the future, if the system ever runs between the Bay Area and Union Station, operating speeds might be reduced to say, 155 mph, 250 kph, or even slower to 125 mph, 200 kph to save money while running between stations but also to limit the stop and speed resumption costs. (It's terrible to think of track quality and maintenance that matches today's road quality and maintenance in the state, and may force speed reductions, or skimping from much earlier in the project's execution.) I'm still not considering any skipping of stops any more than the project administration is, because of the politics of why so many stops have been included since the beginning.
DeleteCHSRA is building a $200 million solar facility, which would at least help during daytime hours.
Delete"HSR an energy hog at speed" in comparison to what, exactly? If you compare to the short-haul airline flights that HSR will replace, HSR looks downright thrifty.
DeleteWith the California electric grid trending rapidly towards solar + battery, that energy will be 100% renewable. Look it up, there's something like 4 GW of battery generation capacity coming online PER YEAR. Right now.
That's why I am convinced that "HSR is an energy hog" is a bullshit argument.
Re: insufficient toilet capacity, I strongly believe the toilets should become a wayside facility. The onboard toilet, and providing ADA accessibility to it, is at the top of a domino cascade of techno-regulatory consequences that stand squarely in the way of level boarding. Toilets belong in stations, even if their maintenance won't be cheap.
Delete@Richard:
DeleteRe "baby bullet" being a harmful throwback, "2024, something improved a bit" and current Caltrain staff having no experience in making/changing time tables:
Maybe it might be possible to leverage all this?
The reason that something improved a bit is HSR.
No matter what competence or lack of competence Cali HSR might have, we can at least be sure that they likely don't have 20 years of experience of just maintaining status quo.
In other words, maybe Cali HSR could come in and talk something about "preparing the Caltrain route for HSR" and call the express trains "mini-HSR" or "pre-HSR" or something like that, and in doing so also stating that for HSR to run optimally the time table needs to be a certain way, more or less just telling Caltrain how to do their job. Maybe I shouldn't really suggest this here, as Caltrain staff might read this and oppose things just to be stubborn, but still.
Even with a 1% success rate chance it might be an avenue worth pursuing?
@Reality Check, @Anonymous and @Clem:
In addition to HSR being more energy efficient than flying, electric trains are in general way more efficient than private cars. Not sure exactly how the comparison is with electric cars, but at least in comparison to combustion engine cars the cars are way way less efficient.
Not 100% sure but I would think that electric trains would be better for the environment than diesel/gasoline cars even if the electricity comes 100% from coal.
@anonymous
Can't remember what proposed final service pattern Cali HSR might have published, but of course some trains can run non stop to both achieve short travel time and less electricity consumption. (Arguably full non-stop trains is a bad idea, it's better if they stop at one station near each end point, I.E. SJ near SF and whichever somewhat local station might be the closest to LA Union station).
Either way, building a long HSR train line without leveraging it with additional stations seems like a really bad idea.
(This is also why I think it's a bad idea that the HSR don't include a station at Las Banos - the only decently sized populated area along the route that won't get a station).
@Reality Check and @Clem:
DeleteTotally agree with Clem that toilets belong at the stations.
Toilets on trains seems to be a thing related to stations with no fare gates, resulting in that toilets at stations having a fee to use the toilet (as the transit agencies / rail/station infrastructure owners doesn't have the task of providing free toilets to the general public), and thus travelers prefer the free toilets on board rather than paying at the station. A solution for this would simply be to have the toilets inside fare gates. There could even be a special "poop fare" only valid for entering the station to go to the toilet (and/or meet-and-greet people - a century ago platform tickets were a thing in some parts of the world). A problem with toilets in general in cities is that they tend to both become homeless shelters and also a place where people go to to use drugs. But that must be the same problem for any place with a roof that the general public has access to, and whatever solutions are done there would apply to toilets too.
80% of the on board tank capacity seems a bit low. Maybe better safe than sorry though? In Sweden it has happened at least once that the class Y2 trains didn't close their toilet when the tank was full, and eventually the train "pooped it's pants", i.e. sewage overflowed and ran out into the passenger compartment.
Btw, isn't it better to just have fixed sewage emptying facilities at all tracks at the end station in SF and also at SF Diridon? I.E. while reversing the trains just hook up a hose to empty sewage, and disconnect the hose when done, rather than driving a sewage truck to the train? Don't know at what rate it would be ok to send sewage into the city sewage system - if there is some rate limitation then there could be a holding tank at each station that trains empty to, and it's content would then empty out into the city sewage system at whichever rate the city sewage system deems good.
@Mia
DeleteThere's really no prospect of adding fare gates, as none of the stations are configured for it at all. Nor would it be worth the expense. Proof-of-payment works fine. Even if fare inspections were less frequent than they are now (I get scanned probably every other trip), it would be fine. Just put up a JCDecaux self-cleaning toilet or something similar.
@Reality Check
Do we know if they could modify the consists at CEMOF (in a reasonable world, real-world COO protestations aside) or if they would actually have to send them back to Utah? Shortening trains seems like a good idea for the foreseeable future, especially if this makes it possible to increase frequency due to lower operating costs (staff and energy) once more are delivered.
Clem: It has never a bullshit argument, and you (should?) know it -- not any more than saying things fall down rather than up or sideways. A comparison is irrelevant; the energy costs are high to travel in the high-speed realm, as aerodynamic resistance rises substantially, a fact known with conventionally powered trains and road vehicles as well. It's not for nothing that the CSHRA faces high energy needs and costs to meet the power requirements to move trains at high speed. Can trains run fast (or far) on batteries as the likes of Laura Friedman apparently believe?
DeleteComparisons, a separate topic, are in favor of electric propulsion, because of much greater efficiency. In fact, at the cost of high energy use, this efficiency is exploited by high-speed railroading to travel practically at speeds far above that of trains and road vehicles that are powered conventionally (by internal combustion). That is the normal comparison, particularly with trains limited to 125 mph or so, but one also says now the first, because exploiting that efficiency means electric trains can be run fast enough to be competitive with air travel at the longer road and shorter air travel distances, and electric high-speed trains are more efficient (in addition to being cleaner, etc.) than air travel despite aerodynamic resistance at the ground level versus at altitude, with which you're no doubt familiar. Thus high-speed trains are routinely compared with air travel, too (at the right distance range or travel times). That's still not negating the obvious high energy use of fast trains, multiple trains at once if the state ever got a serious train system.
@MiaM, the high-speed rail legislation includes a ban on stops between Gilroy and Merced, intended to prevent a stop in Los Banos. Some were politically opposed to people moving there and commuting to and from the Bay Area on high-speed rail.
DeleteThere are so many stops along the Central Valley because that many were agreed to in order to secure political approval, too. Fresno and Bakersfield arguably should get high-speed service along the way, but the other cities are too small. They can benefit from rail, especially replacing air service, notably in winter with fog a hazard then. This should really be a system supporting more than one kind of train and more than one kind of service. Meanwhile, my original thought and prediction stands, that the future could see the state reducing speeds or for that matter, stops ("for improved performance") to reduce power and costs, if the system were ever completed and running at full speed and power.
"there's something like 4 GW of battery generation capacity coming online PER YEAR."
DeleteBatteries don't generate power, they store energy. All of the power to run those trains has to actually be generated somewhere else before going into and out of a battery again.
"Toilets belong in stations, even if their maintenance won't be cheap."
Somehow Target, Walmart, local gas stations, minor league baseball stadiums, bars and restaurants manage to maintain public bathrooms used by many people a day without going bankrupt.
@Onux: Indeed, and it's a lot of power that is needed, with making cuts in performance as a way to reduce the demand in the future, if needed. (See Jane. See Jane run. ...) The figures for the state system being thrown around are impressively substantial, e.g.,
Deletehttps://hsr.ca.gov/wp-content/uploads/2022/09/final_EIR_MerFres_TA3_06C_EnergyUse-A11Y.pdf
versus the size of this, incidentally:
https://hsr.ca.gov/programs/green-practices-sustainability/sustainability/
@Onux: thank you for the important clarification that batteries store energy. The entire California grid peaks at order ~50 GW, so 4 GW of battery (power, not energy, with a duration of 2-3 hours being implicit) coming online every year represents a rapid trend towards utility-scale solar + battery taking over everything else. Turns out the fusion reactor was right there all along, in the sky. The trend towards a 100% carbon-free grid will culminate within ten years, well before the statewide HSR system is completed.
Delete@Anonymous, I still can't fathom your argument that HSR uses too much power. Power is plentiful and green, and getting rapidly more so every year. If you want to hand-wring about excessive energy use or climate change, focus your efforts on internal combustion road transport; HSR is the wrong target.
Clem, I have never said the trains use "too much" power, and don't know why you would imagine that; it doesn't follow from "energy hogs" used properly in context, obviously. They use much power, not "too much" power. (They are traveling very fast, exploiting the efficiency of electric propulsion, to compete with air travel over reasonable distances and travel times, and some want to see more of what France has pioneered with replacement as government policy.) It's no different than exploiting LED efficiency to increase illumination currently installed somewhere and illuminate new places as well, for example. There is the chance of cost reduction in the future by reducing power use (not required to apply to all manner of future services, non-stop or limited-stop versus all-stop, incidentally) and reducing the system's performance from intended super-BART service. The state faces cost rises in the future for a number of things or reasons, a fact independent of and addition to however eventually successful it is at developing the substantial solar and wind resources. There is no hand-wringing, just no surprise by those who are aware if things change later, as with decisions already made as the project has progressed. Many of us would love our own Milan-Rome (albeit without a real Florence analogue here) on trains like these, and electric trains elsewhere, but as with Caltrain, the state hasn't been the best performer and we can see more forms of disappointment or unpleasant surprise to others later.
DeleteOh wow, a legislation specifically targeted at not allowing a station in Los Banos is some weird s**t :O :O
DeleteRe power usage: With regenerative braking, sub stations near passenger stations, and battery storage near said stations, stopping a train shouldn't use more energy than through running without stopping. Sure, some energy is lost partially due to having to use mechanical brakes, partially due to losses between the trains motors and the batteries (and in the batteries), but on the other hand the drag is of course lower as the train slows down to a stop and then accelerates again instead of running at full speed for the same length of track.
In other words I don't think that it would save power to run all-stopping services at a slower max speed than non-stop / limited stop trains.
Btw, I get that both a lot of the California inland and more so it's neighboring states are sparsely populated, but anyway: Why are there so few grid connections to the neighboring states?
anon if you're really losing sleep over 10MW trainsets then I suggest you get to work digging 42 minute chords.
Delete@MiaM, try this, which includes related information. The proposition's language was how Los Banos was excluded as a station site. ("There shall be no station between the Gilroy station and the Merced station.")
Deletehttps://repository.uclawsf.edu/cgi/viewcontent.cgi?article=2279&context=ca_ballot_props
The Sierra Club (which brought us the first carpool "diamond" lanes) and other environmentalists opposed the ideal site to support living in the Central Valley and commuting to and from the southern Bay Area or elsewhere on the train because it would mean "urban sprawl" and loss of farmland, et cetera, in the Central Valley.
There are environmentalists who support high-speed rail and some opposed at least to some potential design elements, just as there are some who oppose solar or wind power because of deleterious effects on the environment, while most environmentalists support it. The great majority of environmentalists and others holds the normal positions on these matters, but occasionally the other side can win.
@Anonymous:
DeleteThanks.
For anyone else following the link, the juicy part is on pdf page 12:
"(d) The total number of stations to be served by high-speed trains for all of
the corridors described in subdivision (b) of Section 2704.04 shall not exceed
24. There shall be no station between the Gilroy station and the Merced
station."
If someone would want to work-around that legislation, a loop hole would be to build quad track between west of Los Banos and the junction between Merced and Fresno, and only have platform tracks on the tracks towards Fresno. Would also require removing another station (if the total is already 24).
The bonus question is if the legislation says anything, and if so what, on connecting to other rail and if so if it says anything about mixed operations? Otherwise it would be easy to just add junctions to add additional stations that technically aren't part of the HSR legislation but still would be used by trans that also run on the HSR line.
Los Banos might be a somewhat obvious place, but the other places would be all the towns/cities between Fresno and Sacramento that has San Joaquin service but won't have a HSR station.
@MiaM, many want the whole project put on the ballot again. Yes, that would be a referendum on the project, possibly to stop it or to curtail it in some way.
DeleteSeparately from that, the legislation now governing the project from could be changed, including the excision or the removal of the ban on stations along the route crossing the Central Valley and the mountains (Pacheco Pass). Removing Gilroy as a stop as part of replacing Pacheco Pass with Altamont Pass and adding a stop at Manteca as well as in the Tri-Valley area would be a great improvement as well but I don't expect it to happen with San Jose and related interests continuing to have political clout. Also, might certain interest say that changing the system is a form of reneging on conditions to be met for the bonds that are issued for this project. (Don't overlook that some might use that as a way to fight an "urban sprawl" Los Banos station in court.) Cynics can add, the project no doubt will fail to meet a number of its requirements, so why not be honest and remove them?
And yes, what really should be here is a full system with more than one kind of train service at different speeds, but it won't happen.
Meanwhile, any Caltrain is better than none, but can't it do better?
It can be argued that a Los Banos station can be funded with other sources other than the 2008 Prop 1A, as it is a bond authorization law, and with it where the bond money can be spent. The 2008 Prop 1A will cease to have any influence once all the $9.95Billion has been spent, along with its ban on a Los Banos station.
DeleteWhat would it take for them to run 3tph, the same as BART?
Delete"What would it take for them to run 3tph, the same as BART?"
DeleteJust deciding to run 3tph, same as BART! That's what.
I believe (my slogan on this one and only this one matter being "don't overthink it!") that it might work nicely: 3tph all-local 30s 15% 6MW; like 2024 BART, but slower and more expensive (Fleet 11 trains; Score 115).
Note 11 trains and crews, not the 14 Caltrain presently doesn't even have available, and 75%(!) fleet utilization.
Note that 11 trains for all-day 20 minute headway service compares not-unfavourably with the 8 trains Caltrain requires for 2tph off-peak.
Is that The Right Thing to do? I don't know! But I'm not at all sure it isn't. I do think it's worth thinking long and hard about, especially as Caltrain's "commuter railroad" ridership simply isn't coming back.
It's all trade-offs. Maybe this is a good one. Maybe!
@Anonymous and @William:
DeleteThanks for the replies.
Waiting out things until the money from the ballot it spent is probably the safest options.
I've been babbling about this previously in the comment sections, but in general HSR mostly covers the same places (albeit differently) between Merced and Bakersfield as the San Joaquin trains does, but between Merced and Sacramento there are way more stations on the San Joaquin line than what the HSR proposal/plan has, and it seems like a good idea to include (equivalents to) all these stations for running local trains.
Taking a few examples from Sweden:
Gothenburg southwards (towards Malmö and Denmark) is a double track line. They seem to run 4TPH local trains stopping at all commuter stations, and 4TPH regional trains stopping at one station in every county (of which some are "Öresundståg" running towards Malmö/Denmark, and some only run locally within the region). In addition to this there are freight trains, but I don't know if they run peak hours.
Gothenburg eastwards has 2TPH that stop at all stations, and 2TPH regional trains. In addition to this there are X2000 trains Gothenburg-Stockholm, iirc every hour, and afaik there are also competitors to the X2000 trains. And also freight trains, although again I don't know if they run rush hour.
The point of bringing up this example is to show that it's possible to mix slower all-station trains with faster limited stop trains and still have 4TPH for each service on a double track railway. Sure, you really have to squeeze in things in the schedule, but still.
Also worth considering is that the rush hour for local and long distance travel aren't always at the same time. IIRC in particular the afternoon rush is at slightly different times. That means that the schedule can be optimized for longer distance limited stop trains at certain times of the day, while at other times the schedule can be optimized for local all-station trains.
@MiaM: It will take so long with this project -- yes, the bond money will all be spent. Of course, better than a Los Banos stop, on a Pacheco Pass route, would be a Manteca stop and transfer station, on an Altamont Pass route.
DeleteAs for your remarks about faster and slower trains, yes, the state should have a true rail system, which would support different trains and kinds, and levels, of service on the Central Valley "trunk route."
I don't get Richard's hatred for the baby bullets. Around a decade ago, these were absolutely the most regularly packed trains that Caltrain operated. Both the 7:59 am and 8:14 am SB departures from SF would often be standing room only. Perhaps you'd wait for a seat on a local train, but everyone standing on those trains would argue that their travel time is too valuable to take a slower train.
DeleteAnd that's why it strikes me as odd that you'd argue that Caltrain provides slower trains when nearly every time someone gets directions on Google Maps, Caltrain's speed is in direct competition with driving.
Re: "I don't get Richard's hatred for the baby bullets."
DeleteMartin, in your world, a rail line representing tens of billions of dollars of public spending exists for exactly two round-trips for you five times a week, 50 (America, fuck yeah!) weeks a year.
The rest of the time it doesn't exist, because your crossover (America, fuck yeah!) SUV is available in your driveway 24/7 for zero-wait-time, no-intermediate-stops, point-to-point (a pint of ice cream! DRIVE!) trips, on demand. Zero friction!
So naturally it is super important that the trains running that one round trip per day, 5/week, 50/year, be as fast as possible, and make as few irrelevant intermediate stops as possible, because it's important that you get back to your home with the garage and the crossover SUV as quickly as possible, because that's how you maximize your quality of life.
The rail line and the trains don't otherwise exist. Buses don't exist. The poors don't exist. Commuter Rail is for Commuters. Commuters gonna commute. Commuter want "baby bullets." That's that commuter trains do. Time is money(*)!
((*) my time, other people's money)
There's no need to be odd (keeping it lenient) about Martin's remark about the Baby Bullets. As an alternative, to be consistent then also state there should be no non-stops or skip-stops for the promised future high-speed rail service, either. Not that the 2:40 and other specified maximum travel times have been destined to be met given "progress" with the system to date. The politicians wanting stops along the way to secure initial progress aren't involved here.
Delete"As an alternative, to be consistent then ..."
Delete... then, but whatabout the Iraq War, then? Surely. Surely!
Martin is happy with one train per direction per hour for his commute, and is happy to arrange his work life around one train per direction per hour. Happy Martin!
Other people may not be so happy about one train per direction per hour. Either because they do not commute to and from their jobs to and from the same stations at the same times every day as Martin does, or because they aren't "commuting" at all. Some non-Martins may be less than happy!
To be non-leniently odd, the world is a place of diversity and contrasts.
I get why Richard doesn't like the "baby bullets", and that's not that they are faster trains, but that they take up space from other trains.
DeleteIf it was possible to schedule them so that they have good interchange with connecting lines, and also have equally good interchanges for the all-stopper trains, then things would be better.
But also, they spent loads of money on improvements, with partial quad tracking and whatnot, and are yet not able to run 3-4TPH slow trains and 1-2 TPH fast trains on the same double track railway.
This relates strongly to that almost no-one have any experience in planning completely new schedules or schedules that differ vastly from previous schedules.
Also, how are they planning on combining Caltrain with HSR trains if they can't get faster and slower services with a decent frequency to work without adding HSR?
P.S. if this had been in Switzerland they might had decided that the Caltrain EMUs and the HSR EMUs should be possible to mix within the same train, and then just connect and disconnect the vehicles in SJ, so a "bullet Caltrain" train and a HSR train can use the same path in the schedule.
@MiaM It's very much not a question of available track capacity, but money. They can't run more service because of (largely self-imposed) high operating costs.
DeleteFor what it's worth, to run a sensible schedule with a mix of slow and fast service (excluding HSR), the existing quad track segments are in the wrong places. If you look at the 2040 business plan, the only place where Caltrain needs a passing segment for their own operation (Redwood City, based on this stringline) is precisely none of the places where four tracks exist today. The only way to add HSR into the mix is to add more four track.
@MiaM, Richard's reaction was another wild swing and miss, Iraq and all. [sigh] There is a place for all kinds of services and for that matter, speeds or there isn't; it's not limited to Caltrain on the Peninsula but also applies to the entire future state high-speed rail system. In practice, nobody yet is eager to proceed with planning on limiting stops and losing Central Valley politician support.
DeleteI'm not counting on the future promised (i.e., very remotely possible) high-speed trains to run at 110 mph or close to it (what curves permit) between stations along the Peninsula route and between San Jose and Gilroy as assumed by some, necessarily. What can be done economically in addition to practically is fine for both Caltrain and the high-speed trains reduced to conventional (slow) speeds in the metro areas and elsewhere it is more developed. (What if cities along the route wanted trains slowed to reduce the noise they make at higher speeds?)
As far as Caltrain service and desired frequency, a real-world measure is nice, especially if one is transferring or switching travel modes. If one train is leaving when you arrive at a station, at any location where you can view the tracks, you want to see the headlight, at least, of the next train. Short of that, you don't want to wait more than, say, 5-10 minutes.
@Anonymous
Delete"It's very much not a question of available track capacity, but money."
More service period is a matter of money yes, but as you said, more service with a mix of locals and baby bullets is a matter of quad track capacity. So Mia is right about the problem with the baby bullets: the overtakes are in the wrong places for a lot of fast-slow service, so insisting on running the baby bullets means non-bullet stations are left with poor levels of service, even if operating costs improve. This reinforces Caltrain mostly being useful for 9-to-5 commutes (and park-and-ride to Giants games) more than all-day casual travel, which requires all stations to have frequent trains. The fix (in addition to fixing operating costs), as you said, is to put an overtake where it actually makes sense (Redwood City).
"The only way to add HSR into the mix is to add more four track."
Even quad-tracking the *entire peninsula* wouldn't solve Salesforce throat capacity (12 total tphpd if it's a terminal), leaving either HSR or Caltrain or both stunted forever. I know this is my thing that I never shut up about, but there is a way to get HSR to San Francisco without permanently strangling Caltrain: HSR should arrive in SF by way of Altamont-Dublin-Oakland-Link21 and not go any further.
And hell, with no HSR on the peninsula you could someday run a peak of *16!* Caltrain tphpd (8 local 8 express; probably 4 go to Dumbarton), all through-running from SF to the east bay, without quad-tracking the hardest sections of the peninsula or anything in SF. We could fully leave ye olde 9-to-5 commuter railroading behind and meet capacity needs through smaller and more frequent trains (with smaller staffs) instead of a few massive consists per hour--and we could even keep our baby bullets.
(Salesforce western throat can handle 20 tphpd with through-running; I am assuming that the Link21 tunnel could handle at least 24 tphpd and that the eastern throat would be better-designed and could also handle 24, so you could terminate 8 HSR from Oakland and through-run 16 regional trains to the Peninsula)
"For what it's worth, to run a sensible schedule with a mix of slow and fast service (excluding HSR), the existing quad track segments are in the wrong places."
DeleteYeah, anybody with a brain has known this and been on the record about this since the early 1990s, back when Caltrain's "Project Ponderosa" added the quadruple tracks in obviously the wrong places.
The bad news is that no people with a brain are now or have ever been Caltrain or CHSRA consultants or staff.
(FYI I fully admit I was wrong a couple decades ago, thinking Hillsdale should be the central overtake/transfer location for the corridor, based on some incorrect train performance guessestimates. Wrong! Redwood City is, rather, the correct central turnback/transfer location, based on more-correct less-wrong data, and clearer thinking.)
"The only way to add HSR into the mix is to add more four track."
The good news is that very very very little is needed! If you have a brain. Which ...
Anyway, just Bayshore – South San Francisco. Easy easy easy easy. This without HSR versus this with 4tph HSR. A little further south through San Bruno might have been nice, but maybe it is not absolutely necessary.
Anyway, all angels with pinheads. HSR service isn't coming to California within our lifetimes, though of course CHSRA spending will continue until the rising oceans make it all moot. The beatings will indeed continue until morale improves.
@Anonymous (1?):
DeleteIn other words the problem is kind of that they bought 7-car trains instead of a mix of a total of twice as many 4 and 3 car trains. Sure, it would cost more staff and whatnot but the maintenance and investment cost would not be much higher with twice as many half length trains.
Also re the business plan: I get that some stations have low passenger numbers, but it seems bad to have a 15-15-30-15-15-30 min service at Menlo Park and a 60 min service at Atherton. Would it be politically impossible to close Atherton?
Also alternating services giving at best a 30 min service (and at worst at 15-45-15-45 service) Milbrae-San Mateo doesn't seem great.
College park and the airport should rather be serviced by a tram (VTA light rail line) SJ Diridon - Santa Clara (and connecting to whereever there already are VTA tracks at each end).
It's also a bit sad that they propose 4 TPH HSR to SF, and terminating 4 THP HSR at SJ Diridon. This on the other hand kind of asks for those trains to continue to Oakland on an improved and electrified railway. It's unfortunate that the total run time would be really long if they would run through Link21 in a loop and continue back to LA. The future will show if users are "kind enough to the trains" for that to be possible without needing cleaning and whatnot. In theory HSR could allocate parts of the trains for travel between different stations in a pattern resulting that during a bay area loop certain cars would be empty of passengers between different stations, allowing some simple cleaning to take place while the train is on the move.
The other super long distance future option would be to combine those 4THP HSR trains with a future electrified and greatly improved Capitol Corridor service, as discussed in the comments of other posts on this blog. If this is done this should be the HSR trains that stops at most stations as it would be bonkers to take this route between the Fresno-LA part of the HSR and Sacramento rather than the direct route through Stockton. For regional travel it could be useful though.
@Anonymous (2): Yeah, Iraq is a bit far too off-topic.
Re noise: I doubt that non-stop electric trains at 110MPH makes more or even an equal level of noise as compared to all-stoppers using the diesel trains that's now being phased out. In other words Caltrain just has to run some noisy diesel Gilroy service until HSR arrives, so HSR arrival would result in less rather than more noise. Unfortunately this is the kind of strategic stubborn obnoxious decisions that seems to be necessary at least in some cases :/
Delete@Bryan Anderson:
A way to get more political support for Link21 is to have some Caltrain and/or HSR services terminate early, and on every time table and whatnot clearly state that Link21 would enable those trains to continue to Salesforce.
IMHO no matter which route HSR takes if Link21 gets built, they should not terminate at Salesforce but rather be trough running through most of the metropolitan bay area. I.E. either SJ-SF-Oakland(-onwards to terminating station and maintenance facilities) or "Altamont"-Oakland-SF-SJ(-onwards to terminating station and maintenance facilites).
More or less this should be the goal for other local/regional services too.
Btw, a bonus question: Would it be possible to elongate the Salesforce platforms when/if the tracks/tunnels eventually get built? Although less than ideal a way to increase capacity would be to allow two trains to use the same platform track at the same time. It doesn't increase capacity that much but at least does something. Also it would allow for possible future super long HSR trains (every other station is at grade or above ground level, and the newly built ones seem to have straight track at each side of them, so would be easy to elongate).
Also, 20TPH with through running at Salesforce seems a bit low, at least if most of the trains are local/regional trains. Sure, for a long while the HSR trains will be a novelty to a bunch of travelers who more or less never have traveled by train before, or only have traveled by the Amtrak "steam trains", but eventually entering and exiting HSR trains will be similar to doing that on your local daily commute train, i.e. about as uncomplicated as taking a seat at the kitchen table in your own home, rather than something even remotely similar to the complicated procedure involved with for example flying (with staff checking that bags are at the right place, seat belts are correctly applied and whatnot).
@MiaM Atherton closed in 2020 and the platforms were demolished shortly thereafter. Presumably all trains that would have stopped there will now stop in Menlo Park instead, giving an even 15 minute headway to that station. As for the skip-stop pattern between San Carlos and Millbrae--unless they build more track capacity, they can't really avoid it. This is the section that is probably the hardest to add more tracks to. Again, in the absence of HSR this is pretty easy to solve (just quad track Hillsdale), but throwing HSR into the mix makes this difficult to schedule for.
Delete@MiaM the platforms at the transit center are already quite long, somewhere in the 300-400 meter range. There is a deferred trainbox extension that could be used to extend one of the platforms to 420 meters, the standard platform length for CAHSR.
Delete@MiaM, trains between Los Angeles and San Francisco should reach San Francisco first at the north end. I say "first" because rather than having all trains simply dwell for a specified time, then go to Los Angeles, there is indeed the possibility of looping through the Bay Area or "threading the needle" with at least some trains. In particular, it's a way to provide San Francisco service off the route that should be taken (Altamont Pass, Tri-Valley stop at Dublin, Dublin Canyon to Oakland, then to San Francisco) or what would more likely be taken (a new Dumbarton route) if the Bay Area were properly served. Instead, the efforts continue to (yes, I correctly say) re-route trains through San Jose.
Delete@Nick:
DeleteThanks for the clarification. Good that they likely end up with an even schedule.
Re platform length: So in other words two Caltrain trains would fit in a row on the same platform track with a CAHSR standard length. If this would be used in practice it would likely increase capacity somewhat as compared to only having one train on each platform. Ideally the time table would be laid out so that the train closest to the end/buffers would be one that needs longer dwell time as compared to the "outer" train. Also this might be a situation where swapping staff might be a good idea. The driver of the outer train just walks to the cab on the inner train nearby, while the conductor (or what you call the other on board staff) of the inner train walks to the outer train. Meanwhile the driver of the inner train and the conductor of the outer train is swapped by other drivers and conductors, and they take a break and then work on another train a few minutes later.
@Anonymous: Side track:
If we anyway are toying with the idea of HSR not going through Gilroy we might aswell suggest going through Martinez rather than Altamont. It would likely not affect travel time that much between LA and SF, and by doing this we would also have a HSR line that can be connected to an easy build/upgraded route to Sacramento. Heck, if we are toying with the idea of altering the HSR plans, we could have a triangle junction at Martinez and scrap the Sacramento-Stockton part of HSR, and have trains Stockton-LA pass through Martinez. This way we end up both with most of the HSR objectives met, and also the Capitol Corridor replaced by/converted to HSR.
Side track re this side track: In any case where passenger trains would be moved to dedicated tracks, the authorities/transit agencies have to make sure to get something in return from the freight railway companies as a "thank you" for them not having to deal with passenger trains.
@MiaM: Altamont Pass is the main east-west travel route in the area, more important now with more commuter growth in recent times to augment what began in the 1970s and 1980s. The realistic alternative would be the Martinez-Sacramento route, but that would not support good SF-LA travel, in addition to presenting challenges with terrain and a water crossing. An alternative would be somehow to run trains between Oakland and eastern Contra Costa County, meaning "Lamorinda"-Walnut Creek, Pleasant Hill, Concord, etc., to provide more metro area travel capability beyond BART's, plus regional and inter-city capability as with Altamont. An eastern Contra Costa County route of course could evolve to a Martinez connection rather than go to Pittsburg or Antioch and onward toward Stockton.
DeleteIt's Altamont providing the east-west (in fact, transcontinental earlier) capability that also is why long as it is, Interstate 230-Interstate 238 is and for decades has been the most meaningful large transportation project in the Bay Area. A conventional rail crossing of the Bay to connect to San Francisco and the Caltrain route along the Peninsula and beyond continues to rise in its importance and arguably is #2 now if not a larger combination project that supports buses and a commercial vehicle alternate route to the Bay Bridge. Conventional rail or BART could be added to it, of course; just obtain and spend more money. For rail that is second to connecting Oakland and San Francisco, of course. (It's not sensible to expect to take right-of-way off the Bay Bridge eighty years after more roadway capacity for cars was needed already. A separate project for rail and/or transit (even including light rail, the way things are going, though likely not streetcars) is more than suitable and overdue, call it "Link 21" or whatever; I'm concerned that it will be badly designed as well as executed and money wasted, as usual, with only contractors and some politicians approving.
That's Interstate 380-Interstate 238 (can't edit) that would complete the east-west route and provide a missing crossing route on the Bay, that is likely never to be built.
DeleteRichard. I think you didn't read my full post. As I mentioned, the Baby Bullet service on the two trains I mentioned was standing room only departing from SF.
DeleteWhat are you talking about when you write: "Martin is happy with one train per direction per hour for his commute, and is happy to arrange his work life around one train per direction per hour. Happy Martin!" If you re-read my post, you'll note that there were TWO trains per hour.
To further clarify, two fastest trains departing standing room only demonstrate that there many people do value their time given the option between frequency or speed, do choose speed.
Taking this further, if we CAN fill two trains with people who want that level of speed, then I think we're onto something. However, you keep insisting that we're better off with more frequent slower trains. So I have no idea what your point is as you didn't understand my original post.
I'm going to go a limb and argue that clockface schedules are way overrated once the time between regional trains goes under 60 mins and certainly under 30 mins for the following:
1) Few people plan their commute WITHOUT using Google Maps. Google Maps always shows driving and transit times, so a shorter transit times are highly desirable to get that initial "sale".
2) People are smart and very few show up at a station without checking their next departure on their phone. If I'm walking, taking bus, a work shuttle, or having a friend drop me off, I'm going to try and coordinate my arrival.
3) Clockface schedules of a train every 10, 15, 20, or 30 mins sound nice, but those only work on departures from San Jose or San Francisco. Anyone other trip, is not going to have even spacing, so you'll be on your phone picking the best departure anyway.
@Martin
DeleteI believe you are correct Richard sometimes places too much emphasis on frequency vs speed - the popularity of the Baby Bullet over the years and the number of high speed Nozomi vs all-stop Kodama trains on the Shinkansen in Japan both suggest the importance of speed. However, you have to consider how frequency affects travel time due to waiting. A train four times an hour means an average wait of 7.5 min, vs 15 min for a train twice an hour. For people looking on Google Maps and such, seeing the next train leaves in 25 min vs 10 min is a bigger deterrent than if the time spent on the train is 67 vs 75 min. This wait penalty the same for all trips, but the benefit of speed is only fully seen on the longest trip - 8 min saved for an SF-SJ passenger is only 4 min saved for a SF-Redwood City traveler, but they both wait an average extra 7.5 min in this scenario.
For this reason clockface scheduling and higher frequency (note they are different) is actually more important the shorter the journey - the opposite of what you argue. If you are going to travel 6 hr somewhere you really don’t care if your train/plane options leave at 7:47, 2:11 and 4:53 - for a journey that long you will plan to leave at one of those times no matter how infrequent or random they are. But for a journey less than 30 min a frequency of only 2 tph means Google might tell you to wait 20 min for a 26 min ride - that situation is when you lose people to driving. Waiting 5 min for a 30 min ride looks more attractive.
Re your 3), it’s not true that clockface schedules only work at the SF and SJ ends. If the train leaves from one end every 15 min, then it will arrive/depart at each station every 15 min - the spacing remains even as long as the stops are the same.
What I think you mean is that a departure from SJ at :00/:15/:30/:45 becomes a departure from Lawrence at :11/:26/:41/:56. It is true that this isn’t as intuitive at :00/:15 etc., but it is still valuable. That :11/:26 etc. pattern would repeat all day long. If that is your home or work station you learn it and can rely on it. It helps plan spontaneous trips (“A client wants to meet in SF, its a 4 min walk to the station, train leaves in 7 min, I won’t have to pay for parking.” or “I can stay to finish the project and take the :56 home instead of the :26.”) as well as commute routines (“Leave house at 7:30, drop kids at school at 7:55, catch 8:11 train” or “Spouse takes kids, leave house by 7:50, catch 8:11 train.”)
But even further, SF and SJ are not necessarily the only stations with the :00/:15 pattern. The :00 departure from SF would/could be a :30 departure from Redwood City, and likewise the :15 becomes the :45. The :00/:15/:30/:45 from SJ becomes :15/:30/:45/:00 from Mountain View.
Now where it really gets interesting is what Richard has been illustrating that the :00 departing SJ could ALSO be the :30 from Redwood City, and that if the :30 from SJ leaves Mtn View at :45 northbound then the :00 from SF leaves Mtn View at :45 southbound as well. At 4 tph you can have up to 7 stations between the ends where there isn’t a northbound or southbound train time, there is just the train time, with the train in either direction leaving within a minute or so of each other. This means all the convenience of a clockface schedule extends to every train journey you take, to any other station (not necessarily on return, but luckily the stations this applies to are mostly high ridership/destination ones, so it could often be both ways). It also means if the bus arrives around this time you can be dropped off and get right on the train with no waiting, while the person getting off that train finds a bus waiting for them, instead of waiting for the bus. So you can take that 7.5 min average wait time and make it effectively zero for many people, shortening trip time even more compared to the “faster” 2 tph option.
Wow, 48 minutes Diridon to Gilroy, on top of however far you come from SF or Silicon Valley to Diridon. And still, commute direction only. - Ben Pease
ReplyDeleteBen, 48 minutes Gilroy-SJ Cahill is actually two whole minutes faster than today's schedule. 4% speed up. Choooo-choooooo!
Delete4 trains per direction per day is as high as it's ever been, plus a whole 4tpdpd higher than has ever been justified.
Plus you get Caltrain's "good" locomotives and the Bombardier cars. An entire fleet luxuriously dedicated to under 600 riders/day (< 150 riders/train/day -- y'all fit in one Bombardier car!)
And, as we all know, there's a far superior solution.
Everybody else ought to be more worried about not-unused Caltrain service north of SJ Cahill being held up and messed up to wait for the scheduled connection (scheduled to be far better than it has ever been!) with the UPRR-hostage diesel trains coming from Gilroy in the morning. Tail wags dog.
Fun to look back at post from six years ago where I argued that Caltrain was over-promising the run time savings of electrification.
ReplyDeleteI think we'll soon see the same phenomenon that took hold after the introduction of the Baby Bullet in 2004: dwell time padding and run times will gradually increase from a "sporty" timetable that was first introduced, because long and unpredictable dwell times are a stubborn problem that Caltrain continually refuses to admit, let alone solve. Even seven megawatts of EMU power won't fix that!
Re (hoisted into a new reply thread): "insufficient toilet capacity, I strongly believe the toilets should become a wayside facility. The onboard toilet, and providing ADA accessibility to it, is at the top of a domino cascade of techno-regulatory consequences that stand squarely in the way of level boarding. Toilets belong in stations, even if their maintenance won't be cheap."
ReplyDeleteYou're 100% pissing into the wind here.
If you look at what actual competent adults do for comparable rolling stock, very recent orders look like this. More on-board toilets, not fewer. (And absolutely yes, "inter-regional" in a tiny county like Switzerland is comparable to "commuter" in a sprawlburb like SF-SJ.)
Maybe SF-SJ is a global outlier! Maybe we're Special Snowflakes! Maybe there's an AI-empowered web-centric wayside-enabled Space-X-rule-rewriting elimination facility solution system just FOR US. Maybe! But ... what are the odds? What does every single historical precedent suggest?
Sure, ADA is certainly a clusterfuck of hijacked good intentions gone largely off the rails, but to blame the syphillitic gift-that-keeps-on-giving of having involved toxic LTK Fucking Engineering "Services" in your rolling stock procurement? You can't blame Richard Nixon for that! (You're just a loser mark who actively enjoys contracting hideous rectal infections, and having others pay to watch your stomach-churning elective medical procedures.)
Anyway, assuming high-floor level boarding ever comes to Caltrain (hint: it won't! Just as HSR will never come to Caltrain within my lifetime or the lifetime of the KISS fleet, because the last 30 years are a perfect predictor of the next 30 years) then the solution, as I've mentioned, is just to buy some high-floor, single-level, toilet-toting unpowered cars from Stadler, who are happy to build any utterly stupid shit (or any sensible shit) for anybody who comes forward with the cold hard CHF. Cheaper than Smart Pipe, I 100% guarantee, regardless of the exchange rate. After all, Stadler just take your money; they don't poison the well for generations, unlike scum of the ilk or LTK or PTG or AECOM.
I know what you think of this idea, but... if Link21 is built mainline, and mainline also steals the BART tri-valley branch and connects to the central valley via Altamont, that would actually be cause to double the fleet and replace all the cab cars on one end with FLIRTy versions as you proposed in the linked thread.
DeleteThis would also solve a lot of bigger problems if HSR does ever get completed, like capacity on the Peninsula and at the terminal in San Francisco, in addition to making LA-SF faster than either Pacheco or Altamont-Dumbarton. (Whatever else we disagree on, we agree on the total irrelevance of San Jose)
@Bryan Anderson: Don't just look at the dimensions of the BART right-of-way, not just at the larger regional and high-speed train dimensions, but look at those overcrossings and review the height requirements overhead for these trains. There's also the chance to diverge slightly beyond Dublin (to the east) to connect better with a separate Dumbarton route from the East Bay flats.
Delete@Richard:
DeleteGood reminder that it will take ages before HSR likely reaches the bay area.
@Bryan:
Other possible reasons for procuring more Caltrain trains would be if the proposed improved passenger route between SJ and Oakland gets electrified and gets a frequent ("local express") service, and/or if Caltrain increases the frequency (after all it's a railway with at least double track all the way, with some sections with more than two tracks, so it should be able to run say a train every 5 minutes without any problems), and/or if we eventually get electrification towards Gilroy and additional intermediate stations (and/or just improved frequency to the current stations). Or possibly even continue electrification and add services all the way to Hollister.
Clem, looks like another Richard comment (which I replied to here) was vanished by the spam filter
DeleteSpam filter has been emptied and legit comments republished.
Delete@Bryan Anderson:
DeleteBtw, re BART tri-valley branch: Would it then be connected to another rail route on the east bay side? Or would you cut the BART route to SJ?
If money and whatnot wasn't a problem, quad tracking that branch with two BART tracks and two higher speed mainline rail tracks would be great.
Honestly the more I think about it the more I'm convinced that it would at least be a good idea to study what the result would be if BART would be re-gauged to standard gauge. As I understand it a HSR train would fit a BART station if there is enough height, but the doors would be about 10 inches too high on a HSR train. A BART train might not fit a HSR station/platform. In other words, at least in theory it would be possible to mix both types of trains on the same tracks (although it's a challenge to combine DC third rail with AC overhead wire for the same track, possible to solve but still) and in situations with delays, blockages or whatnot a HSR train could actually use a BART station platform but would require some sort of manually operated ramps/lifts for ADA compliance. This way BART could run a local/stopping service all the way to Tracy and maybe Stockton and/or Modesto.
Btw if HSR would be routed here, a no-brainer would be to run them Oakland-Link21-SF-SJ and allow local travel Oakland-SJ and SJ-SF (but not SF-Oakland) with regular local transit tickets.
You are talking a zillion fasteners set in concrete, a few million concrete ties, switches replaced for the first time after 40 years (rotten wooden ties), and associated wiring which has finally been renewed. Deep down on this blog (a few years ago) the Dublin extension has come up; and either Clem or Richard rebutted the notion that the I-580 route would in any way be suitable for HSR - clearances, corner radius, train/catenary height, and bridge capacity. Not to mention the decades of I-580 widening and lane shifting to get us what we got. - Even if those were back of the napkin calculations I trust their judgement. There were some decent alternative routes proposed on this blog, a little south (tunnels not much crazier than the Pacheco Pass HSR route). It's too bad that BART extension was built as full BART-to-SF (and ended in the middle of I-580, but there it is. Maybe take over that spur and run it as low speed rail with improvements, or single track with passing tracks to whip through curves a little better. (Still probably needs new bridges). More Bay Area traffic slogs over Dublin Grade and Altamont Pass than slogs down through San Jose/Gilroy. (Every Yosemite trip, every family funeral in Stockton) so some sort of rail route to...somewhere useful is STILL a big need... Also for the larger regauging notion, every station on BART has escalators and elevators...and stairs that are the right height for the current platform height. Ben Pease
DeleteThe BART guideway/ROW loading gauge precludes trains that are much taller than BART’s 10-foot tall trains. So forget about room for pantographs & OCS.
DeleteAdditionally, BART viaducts and bridges were built for BART’s light aluminum (40 ton, IIRC) carriages. One of the challenges with developing its new “Fleet of the Future” cars was staying within the guideway’s strict weight limitations.
Atherton station was permanently closed over 3 years ago.
I'm more thinking about future extensions of BART. In particular Dublin-Tracy(-Stockton?), where BART trains could act as slower trains and HSR/Caltrain style trains could act as faster trains. A budget option would then be to have BART trains run on the same tracks as HSR/Caltrain style trains, and Caltrain/HSR style trains could also run through stations intended for BART trains.
DeleteOf course any existing structures that limits the height would only allow BART trains, and the weight is also a thing for existing structures, but still.
Btw, thinking about this, the Caltrain EMUs have about 2"/5cm lower floor than BART trains. Maybe they were partially thinking about this?
Or more likely they just opted for whichever height the manufacturer had. Seems like there is a tiny 2cm difference between the Caltrain EMUs floor height and the platform height at some S-Bahn systems in Europe.
The lower level of the Caltrain EMUs has a vestibule floor height of 555 mm above top of rail, effectively equal to the standard European platform height of 550 mm.
DeleteRe MiaM (01 Sept 2024), since BART gave up on extending to Livermore (whose citizens didn't actually want BART to come INTO their town, but stay on I-580...so maybe a good outcome until they see the light); ACE Rail proposed to expand their Altamont Pass rail service with a BART connection (on I-580 median, or course, rather than crosswise to BART on the old WAlnut Creek-Danville-Pleasanton ROW now partly bike path and partly office park parking lots). They had proposed short-term improvements, one option had a couple future long tunnels under Altamont and perhaps also Sunol-Niles, to speed up the route. There is a short paragraph here: https://railroads.dot.gov/environment/environmental-reviews/aceforward but the ACE link is broken; this page https://acerail.com/about-us/ starts to mention the existence of the project and a few years ago there were proposals, trainset consideration, etc...and with that, a proposed branch south to Merced which had strong support from the local congressman/legisator.... If ACE creates their own ROW (mixed SP/UP/new) and still kept useful connections to the UP main line, then there would certainly be the opportunity to inter-weave DMUs or EMUs with more of a mid-state high speed rail network...if everyone was broad-minded and the standards accomodated the bigger/faster equipment. That's sorta where thingsa are headed, except that half the people involved won't think big enough, connection-wise and interlining. Hard-luck Stockton (60-90% a redevelopment district before those agencies were abolished) deserves something more than the infrequent San Joaquin and commute-to-San-Jose-only ACE service, but it doesn't need BART (other than hourly or half-hourly standard gauge trains branded as BART). -Ben Pease
Delete@MiaM, BART's plans were to go to Livermore, at the eastern end of the Bay Area there, from the start. It was planned and paid for by ages and wrongly looted as part of paying endlessly for the extension to San Jose with the tremendously over-designed and over-sought subway leading to the still-vaporware Google village around Diridon Station and beyond.
DeleteGoing beyond Livermore to Tracy was never envisioned, though decades after people began living there and beyond, I suspect it's a common thought, expensive as it is. Development beyond Antioch is heading for Tracy, so an outer loop is possible later.
Converting BART to standard gauge means the same size rolling stock, just standard dimension between the wheels. The larger conventional vehicles won't fit in many places, nor overhead electrical power, and structures like the East Bay Viaduct were built for BART-size and -weight equipment. Now, when these structures get in worse state of repair, ... That route is better for passenger rail service toward Fremont, anyway.
Anonymous Earlier (Ben Pease): There's nothing wrong or in any way "insulting" with that freeway route example, and there's no need to satisfy urbanist fetishists with insisting on a literally gross rerouting to divert BART through downtown Pleasanton, then go along the Stanley Blvd. route through downtown Livermore. Now, if there is modern South Bay service, including future high-speed trains, but not ridiculously limited to them, then diverting from the freeway route somewhere east of the existing BART terminus to a common station site for a BART connection and the South Bay (likely Dumbarton) route, which also is the Tri-Valley station any high-speed service should include, that can make good sense.
re Anon 09/01/24 17:03 and I-580 to Livermore, I do like to step off a useful train near the center of the village, but I agree there is value in not digging up and knocking down the village to save it. Most BART stations and their parking lots are not subtle. Following up on my earlier post, a better link to Altamont-and-beyond rail activity is San Joaquin Regional Rail Commission, https://www.sjrrc.com/, who manage ACE, Valley Rail extensions to BART, Merced and Sacramento, Amtrak's San Joaquins, and Thru-way busses. Look under Projects/Valley Rail and Altamont Corridor Vision for their short- and longer-term plans. It takes a lot of work to add a few trains a day, but the San Joaquin aspect will keep them honest, and may provide a good foundation for next steps, and raise expectations.
DeleteBen in SF (Ben Pease): It's too late to fit in high-speed and other conventional trains along Dublin Canyon; don't forget overpasses on the route as well when thinking of loading gauge and the OCS for electrical work. The Tri-Valley, like eastern Contra Costa County, is also part of the Bay Area, precluding high speeds for that reason in addition to limitations in any new tunnels, and Tracy and the area to the Manteca wye site and beyond is also already developed or will be later, meaning its speed realm also will be limited. That's 80 miles through the Bay Area and adjacent area that's developed or that will be, the same as San Jose to Gilroy. I recall an ignorant remark made about tracks in the I-580 median and the inability to reach high speed. Nobody who understands the Bay Area would expect high speed through Dublin Canyon or the Tri-Valley area, with or without the expected Tri-Valley stop. The freeway and older highway route connecting Dublin Canyon and Altamont Pass, not merely providing access in the Tri-Valley and adjacent locations along its way, is fine where it is. It just happens to be a challenge to add rail to it.
DeleteFor real rail there it would be a desire for conventional and high-speed trains performing conventional service there of all kinds, same as service from the South Bay likely involving Dumbarton and Peninsula and City access, too, and would these trains fit, especially when electrified? Not now, notably overhead.
And all this is in a milieu of, as with other transportation matters of all kinds, and with various projects done or merely conceived, MTC and other responsible governments have been failures for decades.
@Clem:
DeleteOh, interesting.
I should had followed the link and not just read the google search summary, to this page on this blog where I god 940mm from. Oops :O
https://caltrain-hsr.blogspot.com/2016/09/emu-brochure.html
Well, then HSR trains will differ enormously from Caltrain trains :( And also if Caltrain trains would ever in a hypothetical future stop at a BART platform there would rather be a need for ramps inside the train than at the station. Might be usable for emergency use but not in regular operations.
@Ben Pease:
Tip: if you click on the "Anonymous" drop-down thing when you comment, you can select "Name/URL" and just enter your name, or whichever name you want (including impersonating all other of us who don't use a blogger/google account).
Seems like the dead url is archived at archive . org :
https://web.archive.org/web/20190306031647/http://www.acerail.com/About/Public-Projects/ACEforward
I just grabbed a randon date that seemed to work, so there might be a newer snapshot from before the page was removed.
Unfortunately it seems like all versions of the fact sheet is missing.
Also the sjrrc url seems to be geoblocked for us outside America. For anyone else in this situation, here is a link to the latest archive . org snapshot:
https://web.archive.org/web/20240627062509/https://www.sjrrc.com/web/20240627062509/https://www.sjrrc.com/
I get why people don't want certain changes, but the two options that would be reasonable in the area would either be to do what the local population seem to oppose, i.e. extend Bart to a place along the current ACE route within the Pleasanton/Livermore area, or extend BART along the along the highway until it reaches a possible interchange station with ACE.
I get that the area far east of Pleasonton/Dublin doesn't really warrant BART trains every 10-20 minutes, but this is a reason for me suggesting a long term future re-gauging of BART to standard gauge, as it then would be easy to build future routes so that you just have to add the BART style third rail and maybe lengthen platforms to extend where "true" BART trains run and where an eBART style train runs. Also this would remove the need for the existing silly extra interchange station between BART and eBART.
Btw it's "only" 46 years since operations ended on the San Ramon Branch. I get that the former right of way is used for other things now, but it would likely be a useful passenger rail route today.
Also, althoug many BART stations have huge parking lots, there is no law-of-nature or similar that mandates that each BART station has to have a large parking lot.
@Anonymous:
My idea with regauging BART is mostly for future lines where it could share the same tracks as other trains, but there might also be existing parts where it might be useful to be able to run other trains. Or maybe not.
@Anonymous 01 September, 2024 18:18: Yes, the problem with going to or through anywhere developed is the challenge that it is already developed. The subject was going from Dublin Canyon and East Dublin-northern Pleasanton along the freeway, where the current station is, to Altamont Pass. It also would make no sense to divert to downtown Pleasanton before taking the rail alignment parallel to Stanley Boulevard to reach Livermore. Nobody would want downtown Pleasanton trashed and the Transportation Powers That Be aren't trusted, anyway. The fetish with urbanism and envisioning revival of downtowns of all sizes has been silly, to keep it kind. People's start and end points aren't in downtowns, as a rule. Those pushing high-speed rail in a vapid or superficial manner discuss downtown-to-downtown, but that's of little value in itself. The appeal it does have is centrality within metro areas. Downtowns aren't normally an origin or destination. San Francisco, working against it as it has in recent times, has remained such a destination. Los Angeles Union Station has a good deal of utility, to be made better if it gets run-through capability.
DeleteFreeway stations are increasingly unpleasant, but sometimes make sense or at least can be accepted as a practical measure. Also, what lies on each side of the freeway and what might later?
@MiaM: It's amusing to consider that the freeway stations with parking lots are particular anathema to the anti-car activist types, Park-and-Ride Plus™. There is also a current mania among some to build housing everywhere possible, and BART parking lots like other parking lots are seen as sites. There is an on-going program to convert station parking lots to housing and build more housing near stations. It's not the same as Brightline (real estate developers with trains, at least for now), but another part of the push for more housing in the state.
Deletehttps://www.bart.gov/about/business/tod
BART should be extended to Livermore and be done with it. eBART itself is silly cheapskate stuff. However, there is an alternative. The key to serious rail service is a modern Altamont Pass rail crossing of the mountains, and once that is done, a standard gauge connection to BART East Dublin-Pleasanton that goes through the new crossing like other trains on the Dumbarton route or elsewhere would be more economical, yes.
The old San Ramon branch is okay through Dublin and San Ramon, but once it reaches Danville it doesn't make sense to use as a route going farther north. The I-680 freeway median and a Danville station at the change would be okay for BART to reach Walnut Creek, but there's no room now. From Danville north the old rail route swings to the west and as a bike trail it was avoided by cyclists because of frequent stops. We cyclists normally have stayed on the road, old route 21 before I-680 came and a classic trunk route for bikes since the 1970s. You can see the rail route here, now a trail, on northern and southern portion pages.
https://www.ebparks.org/sites/default/files/maps/Iron_Horse_north_201200929-web.pdf
What do they give as the reason for the long wait at the end of each run? If it is the time to clean the coaches then that can be achieved in a much shorter time if the cleaners join the train a few stops before the terminus and start collecting trash on the go. When they get to the terminus they can start cleaning whatever is needed starting with the coaches closest to the platform entrance (as passengers are likely to board the first coach that looks to have seats) and finishing the furthest coaces as they travel back to where they joined the train. Passengers are unlikely to pick a seat that obviously needs cleaning even if they do pick a coach that has yet to be cleaned. Should easily be able to get turnnround down to ten minutes.
ReplyDeleteJohn,
DeleteIt's just that American trains get tired from carrying obese people and need to take regular long naps.
Moreover there is no Shinansen-approaching-Tokyo or ICE-approaching-Frankfurt cleaning. This is Caltrain, not Japanese youtubes.
Also:
1. Caltrain doesn't truly believe their own schedules, so, on top of the 15% in-service padding built into the public schedules, they throw in another 20% unadvertised padding in the form of glacial turnbacks. This way a super super late train ("late" as in 25 minutes late on a on 80 minute already-15%-padded run, or 15 minutes on top of 60 minutes) can still return in service on time for the return journey.
Of course padding to provide timetable stability is absolutely necessary, and extra slack is a super good idea when introducting new service; but there's "stable in the face of normal operational perturbations" and then there's "we don't really believe it." We'll see!
2. American Olde Tyme Commuter Railroading practices and regulation are insane, so what would be a technical minimum reversal and return to service for a KISS anywhere in Europe becomes closer to 10 minutes here, because steam trains. I don't know the Caltrain KISS operating details, but FRA mandated brake tests and FRA PTC tests and the world's shittiest hombrew PTC system and just the presence of god-damned conductors mean things aren't going to be a lot better in an Americanized KISS than in behind the wheel of a large God-fearing American F-40 diesel locomotive.
Now this doesn't mean they can't reverse in 10 minutes in a pinch (and as I've mentioned 10 minute turnbacks were briefly scheduled on a couple trains a day back in the 2000s); but it does mean you can't honestly schedule a 10 minute turnback and hope to do it in 4 minutes in a pinch.
Sadly, under the hood, Caltrain remains a steam train outfit, but now with higher infrstructure maintenance costs and higher energy costs (rolling stock maintenance costs TBD, but Clem has good reasons to believe that' can't really change much.) Just like the "quiet Swiss trains" promised look cool, but still BLOW THE SAME APPALLING HORNS and RING THE SAME STUPID BELLS to make sure you know the boat isn't being rocked much at all. Ahhh ....
3. Union agreements. Trains aren't the only things that get sleepy after an hour in service. Conductors gotta conduct!
Side track: What are the turn around times for for example Bart, Muni, VTA and similar? BART has afaik even longer travel times between the end stations for at least some of it's lines.
DeleteTBH if the trains really need more cleaning than just emptying full garbage bags, then maybe future trains should have a water proof interior and a built in pressure washing system that just pressure washes everything at the end stations... This is of course not a serious proposal, but somewhat more serious would be to allow staff to spray water on passenger who drop garbage on the floor (unless it's by accident, or if the passenger is ill or whatnot). Kind of like what some people do to kittens if they misbehave.
Staff needing brakes can be solved by rotating staff, IIRC it's called stepping back on for example London Underground, where each set of staff hands off their trains to other staff, and then after their break takes over another train.
Re schedules:
A great thing about having a higher frequency is that delays that affect each train approximately the same doesn't matter for trains in the other directions, as you can just treat the first delayed train as a missing train in the other direction and then send each delayed train in the time table slot for the next train in the opposite direction. There will be a missing train every now and then but most trains will run and start at the correct time.
P.S. the brake and PTC tests would not need to happen if Link 21 gets build! Well they would have to happen at some place, but not in SF. Either they would happen approximately at Oakland or more preferably only somewhere in San Jose (i.e. run a service that looks like an inverted drop). Sure, we are in the land of dreamers now, but still.
The brake and PTC tests when switching ends are short if the train hasn't been turned off.
DeleteThe Class II brake test is a set and release with a conductor at the rear end, takes 10 seconds
For PTC, the engineer just has to login, enter their PIN, and confirm the track and direction.
The biggest obstacles to fast turnarounds are getting a signal from a dispatcher who's dealing with another problem and the crew needing their smoke break
@Anonymous:
DeleteThanks for the information.
Re dispatcher: How much, if anything, is automatic or near-automatic? I get that end stations and other stations where trains turn around are harder to add automation to, but it can't be impossible.
As an example for most (maybe all by now?) stations of the main/major part of the rail network in Sweden a pre defined train plan more or less dispatch things automatically. I don't know how much has to be done by hand in a normal situation, but in general there is at least an automatic suggestion that a train with a specific train number should be routed to a specific platform.
(Mistakes has happened and those tend to be systematic, i.e. the same train got routed to the wrong platform at Södertälje Syd Övre two days in a row when the train plan incorrectly stated that that train should be routed to a platform with no access to the route the train was destined for, so some cumbersome reversing operation had to take place two days in a row)
Not 100% sure but I would think that turning trains, especially at stations with more than a few tracks, are done manually though. If this had applied to Caltrain there would likely had been manual operations at SF and SJ Diridon, but turning around at Tamien would likely be automatic since you kind of only can do it one way.
Staff has again mentioned there continue to be occasional problems where trains won’t properly initialize or handshake to the PTC system before they can begin their run. The associated delays seemed to be particular or more pronounced in Gilroy the first time they spoke about it. Now it sounds like it also might be happening with EMUs. In this case it often (always?) requires a several minutes-long PTC system reboot to cure.
DeleteJust a reminder that normal Stadler KISS trains come pre-equipped fully-functional, standard, interoperable ETCS/ERTMS PTC.
DeleteNormal Stadler trains ordered by normal customers, that is.
Breaking: KISSes sold in Europe use standard European train control.
DeleteAs if what Caltrain needed was to be the first to US railroad to implement and seek FRA approval for ETCS.
Wabtec's i-ETMS is as close to normal as it gets for the US.
Brightline West applied for and got FRA approval to use ERTMS. The entire process took only a few months.
DeleteToronto is going ETCS. (And level boarding, duh.)
DeleteThose famous members of the EU Australia, New Zealand and the UK have gone ETCS.
Even that famous set of rent-seeking clowns at CHSRA, who are so smart they decided to specify their own globally-unique high speed turnouts because Japs and Yurpeens don't know as much as they do about track, say they're going ETCS.
CBOSS was a choice, freely made, by Caltrain's consultants, Caltrain's staff and Caltrain's executive director, for massive fraud and 100% guaranteed failure. A choice. A choice with clear consequences, if not for them. (Aside from the $300 million tax dollars somebody ended up pocketing.)
Whatever non-working non-interoperable freight shit some local monopolist ended up saddling us up in the fallout of that choice is again a choice, freely made, by Caltrain's consultants and Caltrain's world-class staff, to continue down their unswerving path of pretending to be a Real Murkan mine-to-port coal-hauling railroad in the service of Real Murkan Railroaders, and not an operation that delivers reliable, frequent and cost-effective passenger rail service to the people who paid for exactly that.
Brightline got a waiver for Buy America to import the ETCS components but the system does not have FRA type approval yet, thats a whole other ball game.
DeleteAnd the Brightline West waiver was based on unavailability of an equivalent American supplier. A Caltrain waiver request wouldn’t have gone anywhere because the track speeds on the Peninsula do not preclude using an existing signaling technology that meets Buy America requirements, no matter how glitchy it may be
DeleteCertification process of BL PTC requires only the following:
Delete"Design documentation, testing, and submission of a PTC Safety Plan and associated HSR-125 document will be required to obtain PTC certification and approval to operate. The operational experience of ERTMS across the European high-speed rail network will provide operational safety and reliability data to support the PTC Safety Plan and HSR-125 document."
"Caltrain waiver request wouldn’t have gone anywhere because the track speeds on the Peninsula do not preclude using an existing signaling technology that meets Buy America requirements"
Caltrain will be part of the CAHSR system -- a system that will be using ERTMS. It is absurd to suggest FRA would have denied the waiver.
Re ERTMS:
DeleteCaltrain could argue that they want the same system as Cali HSR opted for. Anything else would mean that Cali HSR trains need to have two systems, and/or that the Caltrain route need to be fitted with two systems.
It would be nice if the Caltrain route in a distant future ends up only having ERTMS and Caltrain forcing freight operators to use locos with ERTMS, and Caltrain could make a few AEM-7 locos available. Just connect the electric loco to the train as-is at around Diridon. Sure, freight wagons are usually more noisy than passenger trains, but not having a diesel loco blasting at full speed would reduce noise and obviously also pollution along the route.
(Don't know if it's okey to haul the diesel locos turned off or if they have to be on and at idle - some locos need to have fans in their electric motors running in order to keep dust from ending up in the wrong place - I remember that some locos that have DC motors need this or for long transports you can take out the brushes (or whatever it's called, the things that transfer electricity to the rotor in a DC motor). I think this for example applies to the Rc locos and thus likely also to the AEM-7/ALP-44 - I wounder if they thought about this when the two units Caltrain has was moved to Caltrain?).
Btw, sorry if I've already written about this, but here we go (again):
ERTMS/ETCS has taken a bunch of inspiration from the Swedish ATC system, which thanks to being one of the later train protection/control systems had a lot of things already figured out almost a half century ago.
A fun fact about the Swedish ATC is that it was almost a skunk works project, except that it had approval from the boss. In order to keep things as cheap as possible it uses a regular CB radio crystal for the transmitter in the loco that is picked up and used as power in the baliser (trackside equipment), and the return signal from the baliser to the loco uses a PAL color subcarrier frequency crystal, which also was readily available and cheap :) The shortest allowable time for a baliser to transmit it's information is so short that it wasn't possible to have any micro processor in the baliser (as just initializing it would take too much time), so it just uses some simple logic chips and transmit a handful of bits (ones and zeros) to the loco.
Also, while ETCS/ERTMS is for the most part great, there has been some tethering problems, and unfortunately it seems like certain software updates can't easily be rolled back. In Sweden the traffic on a higher speed line in the northern parts was stopped for iirc a few days due to a problem with a software update (IIRC some safety problem with level crossings). Unfortunately lines far away from the biggest cities are kind of treated as test beds for ERTMS. The worst case was probably "Västerdalsbanan" which got the "ERTMS Regional" system. It was down so often that the passenger trains on that line barely ran, and eventually that traffic was scrapped... (On the other hand it was one of the diesel passenger routes in Sweden that "don't work" - for some reasons there are only two transit agencies that are able to run diesel passenger services without much problems - everyone else either has all sorts of problems with the diesel trains or they only run electric trains :O )
@Drunk Engineer CAHSR also needs IETMS to work in Metrolink territory.
Delete@MiaM Caltrain has no operable electric locomotives at this time. Of the 2 old AEM7s they bought, one was just for spare parts and the other, while nicely repainted in Caltrain livery, has never been needed or run under its own power.
DeleteFor Burbank-LAUS segment the plan (as indicated in EIR) is add two new separate electric tracks alongside the existing FRA-compliant tracks. Metrolink does not seem interested in running any kind of electric train or shared service. At best LAUS-Anaheim might have shared track someday (who knows)....but insisting Caltrain has to use an antiquated and unreliable PTC due to possibility of HSR-Anaheim is really getting into tail-wagging-the-dog territory.
DeleteAnother tidbit of relevance to Caltrain's PTC decision: PTC Interoperability Agreement with UP
DeleteNot a factor for the EMUs since they won't be on any UP tracks ever, but Caltrain was handcuffed into buying into the Wabtec ecosystem for Gilroy diesel service
"Caltrain was handcuffed into buying into the Wabtec ecosystem for Gilroy diesel service"
DeleteAnd Brer Rabbit sure hated being thrown into that briar patch!
Look, Caltrain (ie Caltrain's rent-seeking perma-temp consultants) choose to enter into such agreements. Just like Caltrain's rent-seeking perma-temp consultants chose to incinerate, literally incinerate, $300 million of your tax dollars on CBOSS. No guns to heads, just fingers in the till. Just cold rational business calculation of what will result in the highest public-private cash transfer ... by the people working "for" the public.
Also a reminder: Gilroy "ridership" = ~100 per train. Stick a fork in it. Nobody gets "handcuffed" by such a nothing, such a demographic and economic zero, unless they really deeply enjoy the entire Handcuff Experience.
Metrolink should be ashamed of not being interested in running electric services.
DeleteI wonder if Metrolink deliverately does as little as possible in the hope that the olympic games will pay for various upgrades?
Re Gilroy:
The amazing (irony) BEMUs will have to run on UP tracks though.
But also: 100 persons per train is 2-3 buses, and IIRC there are three trains in the morning and three in the eveining. That is at least 6-9 buses that would otherwise risk getting stuck in traffic or need to transfer in SJ. IMHO this is better use of a railway than likely almost all of the Amtrak network outside the NEC...
while nicely repainted in Caltrain livery
DeleteHo, ho, ho. Look here! It's a blast from the past.
Although, one must say, the higher-speed example was and is nice viewing, right, MiaM?
Yes ... Yes ...
Just to spell out my reply to Clem's comment "The transformative effect of regular 15-minute takt on connecting busses and shuttles would be awesome"
ReplyDeleteI made that "All true! Also if one ", here's one way that could pan out, with zero effort:
Redwood City
SB :06 :21 :36 :51
NB :05 :21 :35 :51
Palo Alto
SB :13 :28 :43 :58
NB :13 :28 :43 :58
San Antonio
SB :05 :20 :35 :50
NB :06 :21 :36 :51
Sunnyvale
SB :12 :27 :42 :57
NB :14 :29 :44 :59
Santa Clara
SB :05 :20 :35 :50
NB :06 :21 :36 :51
And if you look at the graph, you'll see that by "good luck" (ie my same offsetting of all the northbound departure times to optimize this) Hillsdale (:14/:15 :27/:27 :44/:45 :57/:57) and Millbrae (just twice per hour :05/:06 :35:/:46) also get "lucky" with nearly-symmetry minute bidirectional train stops.
Just imagine how this (and it's just fortuitous non-engineered tip of the iceberg) might have played out in a less shitty world! (Also, imagine that you're a consultancy sucking down tens of millions of dollars, and you didn't notice any of this. Just imagine that. America's Finest!)
Fuck. Fucked up the fucking HTML yet again. blogspot.com is so refreshing old skoool.
DeleteJust to spell out my reply to Clem's comment "The transformative effect of regular 15-minute takt on connecting busses and shuttles would be awesome"
I made that "All true! Also if one shifts either the entire northbound or entire southbound schedule a couple minutes, the glimmers of hints of takt "symmetry points" vaguely start to appear at all of Redwood City, Palo Alto, San Antonio (irrelevant), Sunnyvale and Santa Clara. That's due of course to a lucky rough coincidence of 7.5-ish minutes stop-to-stop between those stations. Imagine what could be done if anybody with a clue were involved in any way in deliberately engineering such "coincidences"!", here's one way that could pan out, with zero effort:
Redwood City
SB :06 :21 :36 :51
NB :05 :21 :35 :51
Palo Alto
SB :13 :28 :43 :58
NB :13 :28 :43 :58
San Antonio
SB :05 :20 :35 :50
NB :06 :21 :36 :51
Sunnyvale
SB :12 :27 :42 :57
NB :14 :29 :44 :59
Santa Clara
SB :05 :20 :35 :50
NB :06 :21 :36 :51
And if you look at the graph, by "good luck" (ie my same offsetting of all the northbound departure times to optimize this) Hillsdale (:14/:15 :27/:27 :44/:45 :57/:57) and Millbrae (just twice per hour :05/:06 :35:/:46) also get lucky with nearly-symmetry minute bidirectional train stops.
Just imagine how this (and it's just fortuitous non-engineered tip of the iceberg) might have played out in a less shitty world!
@Richard:
DeleteGreat suggestion for an inproved schedule!
The bonus question is if it's possible to have the VTA light rail do something similar since it interacts with Caltrain (and Bart) at multiple places (and there is some incentives to also have a good interchange between lines where lines divide, although that is perhaps less important).
(Admittedly I don't know much about Muni frequency, but I assume that at least under some sort of normal situation (i.e. returning to pre pandemic ridership) the frequency would be good enough that there isn't any need for timed transfers - and also to me it seems like Muni kind of only need to time transfers at their southern ends that interact with Caltrain and/or BART, and otherwise just need to spread out the lines to have an even frequency on the shared sections)
@MiaM
DeleteUnfortunately VTA light rail is very poor quality and the orange line that meets Caltrain at Mtn View is damn near useless; take a moment to mentally prepare yourself for some of the daily ridership numbers, especially on the western orange line: https://x.com/QuadMet/status/1823609892399407515
Muni has some high-frequency bus routes where you don't need to think about the schedule, but many of the routes that serve Caltrain stations are not frequent. The two lines serving 22nd Street peak at 15-minute headways and sometimes feel almost as if they're scheduled to just miss every train.
For VTA, Caltrain, and BART, the most important part is to run all systems at the same tph. Whether it's hourly, every 30 mins, 20 mins, 15, etc the agencies MUST agree on keep the SAME tph at the SAME hours.
DeleteI'd argue that if all systems are running at 5tph or higher, then it's a good problem to have, and we don't have to worry as much about transfers.
If Caltrain is doing 4tph during rush hour, then BART should also do 4tph on the Millbrae line during rush hour.
If BART is doing 3 tphs during the day, then Caltrain should also run 3tph mid-day.
Once that's established, then Caltrain can work its schedule so that NB and SB departures at Mountain View and Millbrae align the best they can.
The ideal solution is for all agencies to run 3 tph all day, but in the interim BART could run 6 tph to Millbrae. A 10 / 30 min takt is easier to schedule than 20 / 30 min.
DeletePlans for private San Francisco-Los Angeles overnight sleeping car service revived: https://www.trains.com/trn/news-reviews/news-wire/plans-for-private-san-francisco-los-angeles-overnight-sleeping-car-service-revived/
ReplyDeleteThis company has signed an MOU with Union Pacific and is trying to get a sleeper train between SF 4th and King and LA Union Station in 2025.
Anonymous the First this month: that's a risky venture being undertaken, a sleeper service of any kind. Also, it's a very, very slow route, and improvements sought now won't change that. In fact, it's amusing to see "improvements" as modified later for the Coast Corridor portion of the route. (LOSSAN, with plenty of curves itself south of San Luis Obispo, is separate.)
Deletehttps://railroads.dot.gov/sites/fra.dot.gov/files/fra_net/15521/Chapter%202.0_Alternatives.pdf
There has been related interest on something more conventional, a "Coast Daylight" SF-LA daytime run that might see interest particularly in tourist season. It should be done using nicer Venture cars like Brightline's if possible rather than the already-fine new related equipment used by California and other states. (Not a tilting train to exploit part of the route, or other exotics)
What's needed is a real dining car or at least a bistro car, if not both, rather than mere places aboard the trains with self-service vending. The bistro car on the Talgo Series VI used on the Cascades run, designed by Vergara with a Northwest map on the ceiling, is a great example, could be a model for a bistro on the Central Coast run with brighter "California" "daylight" colors and a map of the related portion of California on the ceiling. Maybe add a light for the train's current position to those for the communities that come on during darker periods including in the tunnels.
The train could continue to San Diego, incidentally. Why not?
Note the state itself could try such a route to see if it's worth it.
The slowness arguably makes the sleeper service more viable than a daytime train. We're gonna have to see if they can get agreements with Caltrain and Metrolink to use their tracks. It's not gonna be super competitive with flying but I think there's real potential here.
DeleteSide track:
DeleteWhy aren't there any passenger trains on the existing freight railway between Bakersfield and LA?
Looking at the time tables, the Coast Starlight and San Joaquin plus bus Bakersfield-LA seems to take about the same time Oakland-LA, and that bus route takes hours. Is there too little track capacity and/or would rail be even slower than bus?
Approx 12 hr route time would be good for sleeper trains though.
The question is if it's really a good business idea to start a service that will die almost immediately after the HSR will be in operation LA-SF as the HSR is supposed to take about 3hr...
Bakersfield to Los Angeles took 5 hours in 1970 (http://www.streamlinerschedules.com/concourse/track7/sanjoaqdaylt197104.html). Compare to 2.5 hours on today's bus.
DeleteI agree the additional time could be acceptable for an overnight train. The state would have to come up with some sort of deal/bribe to get UP to allow a daily pair of passenger trains. Honestly Caltrans should have already secured passenger rights since state funds have been used to double track sections of the route in the past.
@John: I'm thinking of something less risky and dreamy, yes: If a Coast Daylight state or private train were introduced, try extending service to a cheaper overnight run in each direction, too, if turn-around time were kept short, with cheap coach fare. (The Siemens Venture coach "reclining" is something critics don't like.) Next step might be sleeping pods in place of a bistro or dining car. If there's interest, substitute more cars with the pods for coaches.
Delete@MiaM: Why aren't there any passenger trains on the existing freight railway between Bakersfield and LA?
DeleteSouthern Pacific banned passenger trains on its tracks through the Tehachapi Pass after handing off those services to Amtrak in 1971. Current owner Union Pacific has continued the ban. UP and BNSF make a lot of money moving freight (as many as 40 long trains per day) through that line, it's pretty much at capacity.
@MiaM: The Tehachapi route takes hours, as @AG already has noted, and is a very busy freight route, so is not normally available for passenger service. Any service would be fit among freight trains, meaning slow trains, on a slow route. UP does not like the idea even of an overnight train, and has refused that specific request, too. UP is generally hostile to other uses than UP freight on or anywhere near its tracks, and considers rail over Tehachapi useless compared to driving time. Government assertion of passenger train rights in the USA is weak. On the rare occasions an Amtrak train has used the route, it has been a treat in particular for those who understand the details including what good luck it is.
DeleteIt's too bad because the real need and key for state rail travel is connecting Los Angeles and the San Joaquin Valley, i.e., to so much of the state north of L.A..
You might be interested to know that, FYI, the high-speed project will run trains along the same route, albeit on its own right-of-way and much faster, instead of the better alternative Grapevine or Tejon Pass route. The Tehachapi route is used to connect to Palmdale in the desert adjacent to the greater L.A. area, as the old trains did and conventional Amtrak or state trains would do now.
https://hsr.ca.gov/wp-content/uploads/2021/08/BP-Press-release-PHOTO.png
Thanks all for the replies.
DeleteI assume the reason for the route decision is a combination of being able to buy land along the existing railway more easily than new land, and to have a station in Palmdale?
@MiaM: The decision to use the Tehachapi route, that is a dogleg with two mountain crossings rather than one to get to the Los Angeles area, rather than the better Tejon or Grapevine route, was indeed to have a station at Palmdale.
DeleteIt was a political decision, as were the number of stations along the way and worst, the choice of Pacheco Pass over than Altamont Pass. All three were political decisions and all three are fundamental defects in the routing and design of the system essentially from its beginning. (There was support for Altamont, too, in earlier times.)
HSR has been political in the USA, as you may have noticed now with latecomers of various kinds who are very unrealistic, all over this nation (USA) including Moulton in New England, Kunz and his organization, and the crowd in Washington state that views "Hyperloop" seriously, with contractors ready to support that, as an alternative for its own Ultra High-Speed Ground Transportation scheme (of governments, contractors, Microsoft, and others). Nobody gets it sensible. Never mind incorporating it into a larger state rail system in California government's case, for example.
MiaM, politics with the high-speed project is interesting insofar as it tells more of a tale about politics in the Bay Area. The high-speed rail authority's past work has included using a route finding product, Quantm (now part of Trimble), for the northern and southern mountain crossings. Some routing from that use was applied to the segment between Bakersfield and Palmdale (the Tehachapi crossing). Also, one solution identified then as of notable interest for the Grapevine route was used in a work by Clem comparing the the two mountain crossings, "The Truth About Tejon."
DeleteWhere it also is interesting is in the Bay Area. Pacheco Pass crossing routes were examined using Quantm, and also routes farther north by Morgan Hill and San Jose. These were compared to a reference case that actually was on the minds of some as a solution if nothing else were to be known, a base tunnel east of San Jose, 31 miles (50 km) long, directly into the San Joaquin Valley. (!) Now, if you thought that base tunnel was strange, that study did not include the Altamont Pass route as a northern mountain crossing at all. Its exclusion was blatantly political.
While the choice of a "Palmdale" station might have been political, it certainly has many non-political decisions, such as:
Delete1) Enabling a faster connection between SF Bay Area and Las Vegas via the shorter, high-dessert corridor.
2) Breaking up the Bakersfield to LA segment into smaller sections that are more easily funded. A long 38-mile tunnel, will would be on the order the Brenner tunnel in length which will have taken over 30 years between planning and construction.completion. Not the best reason, but shorter tunnels come with less risk.
It may also swing some of the politicians in the LA area into stronger support for CAHSR when they see an actual commitment to connecting up with their area.
Delete@John Ramsbottom, a Grapevine route would also be such a commitment. (Union Station was always in the plan.) However, there were some big boosters of Palmdale as a station site and of the Antelope Valley.
Delete@Martin, who but the rare fan-hobbyist would want to travel between the Bay Area and Las Vegas by train? Palmdale is understandable in being "Tracy South," extension of the commuter shed as well as housing affordability and crowd-fleeing site of early interest, decades ago, in So-Cal as Tracy was for Bay Area people.
The Palmdale route choice does involve two mountain crossings, not one that the Grapevine or Tejon route has, and that's where the work could rationally be divided. What's more of interest to me is not part of the plan, but as elsewhere, to view this train project as part of a state system. Any improved crossing between the L.A. area and Palmdale ideally should serve all regional traffic, not only high-speed trains but such trains fitting into an improved metro extension of an electrified metro system. (That's true for Altamont Pass as well in Nor-Cal and even without high-speed rail up north the improved east-west travel is what is key there.)
It has been written before: If we normally traveled by train instead of by air and on the road for longer distances, there would already be a Grapevine base tunnel for trains, or work underway. That's 34-35 miles, meaning right at the top of world-class tunnel length.
Reducing tunneling? The Authority had a nice session at which they established that minimizing tunnels would be desirable. The Quantm study in the early 2000s found a Pacheco Pass route with a tunneling total of 5.2 miles and longest tunnel of 1.5 miles. However, the approved plan is for a very long tunnel instead, 13.5 miles, plus another 1.5 mile tunnel. That's a high-speed (big-bore) tunnel commitment contractors will love. The Authority also had on its minds a base tunnel near the station at San Jose, of world-class length. ("Because of time and resource constraints, the previous northern alignment studies in the screening evaluation had assumed that the crossing needed to be completely in tunnel because of the difficult and remote terrain. As a result, the only alignment considered included a 31.9-mile long tunnel through the mountain crossing.")
@Anonymous
DeleteYou don't need a world-class 35-mile base tunnel for Tejon/Grapevine. Our host long ago figured out that it can be done with 25-27 total miles of tunnel, none of which is longer than 6.3 miles, removing the need for ultra-expensive emergency egress solutions that are needed in Gotthard-length base tunnels.
http://www.tillier.net/stuff/hsr/truth_about_tejon.pdf
@Bryan Anderson: I'm not saying we need one, or would need one, but again, if we traveled primarily by rail for longer-distance, or "inter-city," trips, we'd have one already or be working on one, same as elsewhere now. It is true that we would have ability now with the likes of Quantm and suitable budgets for a different culture also to cross at less than 3.5% gradient, which leads to more tunneling as gradient is lowered (and shown with studies in California). There might be a trade-off to improve on what was studied, such as the oft-used 2.5%.
DeleteThe Quantm study done with the Authority in the early 2000s included the Grapevine or Tejon route (didn't dare study Altamont Pass and Dumbarton routes, though, politically?). Read Clem's "The Truth About Tejon," including more about politics, and from the Quantm study, there's the following about one of the Tejon, or Grapevine routes found (3.5% max gradient; a steeper route has less tunneling). A desire was to minimize tunneling and preferred, keep individual tunnel lengths to 6.0 miles or less, and it was realized with this route. While much was considered west of I-5 itself, that's not the only place to look:
In the process of studying the possible options to minimize fault crossings in tunnel another possible solution was identified. This alignment option follows the general I-5 alignment up to the San Andreas Fault zone crossing and then proceeds to the east of the existing corridor. Using a 3.5% maximum grade, this particular alignment option allows for an at-grade crossing of the San Andreas Fault zone and an at grade or trenched crossing of the Garlock Fault zone. This alignment would require extensive construction in the floodplain area surrounding Casta[i]c Lake. The potential environmental impacts of this alignment option need to be further studied to assess the viability. This alignment alternative was not
identified in the previous studies of this corridor.
"[... A]n alignment option [...] runs to the east of the I-5 allowing for at-grade crossings of both major fault zones. While this still crosses two faults, it crosses the faults at grade while keeping the tunnel lengths within the 6 miles maximum and maintaining the 3.5% maximum grade. The alignment construction cost is also the lowest of those identified for the I-5 corridor [... .]"
(A total of 17.7 miles of tunneling, longest tunnel length 6.0 miles)
@Anonymous (02 sep):
DeleteWell, all infrastructure decisions on this scale are political more or less everywhere around the world. (The major difference between different places is probably democracy v.s. dictatorship).
It seems though that in USA voters are more keen on voting based on politicians and parties based on their current stance on a few near future decisions, rather than ideology and whatnot?
We've discussed the Altamont v.s. Pacheco decisions over and over. My opinion is that it's actually better that they didn't opt for the best option, as that makes it more likely that both will eventually happen, as the cost-benefit for building Altamont if Pacheco already exists is way higher than building Pacheco if Altamont already exists. Also my opinion is that the bay area needs three good crossings. Pacheco, Altamont and Oakland - Walnut Creek/Martinez, where each of these would be the default route from the bay area to LA, Stockton and Scaramento respective.
@Anonymous (12 sep):
Although traveling Bay Area - Las Vegas via HSR with a future high dessert corridor connection would probably not be that attractive, there would likely be a market for travel between Las Vegas and intermediate stations on the Cali HSR route, like say Bakersfield-Las Vegas or so.
Also as long as HSR don't get some weird security theater (like afaik the long distance trains in Canada has) people will realize how comfortable trains are as compared to the process at airports, and might prefer a longer travel time if that means they can avoid airports. Like imagine being able to just grab your tooth paste if you are going for a week end in Vegas, instead of having to buy a special smaller size tooth paste tube in order to comply with air travel regulations.
Altamont always has been the correct decision and it never will end until not only regional service goes through a modern Altamont Pass route (involving Dumbarton as things go now) but inter-city south of the Bay Area-Sacramento Nor-Cal core, too. I am unsure that Altamont will ever be developed properly, that instead all we may get is increased train service crawling through the current Altamont Pass route and through Niles Canyon. That's again, crawling. Any new Dumbarton service itself is iffy, though I believe some may happen eventually.
DeleteAnd no, "We didn't choose Panoche Pass, instead" doesn't help the lost Pacheco Pass case, re-routing high-speed rail to satisfy big desires of little people with a little city complex in San Jose. (Dumbarton is far south enough.) Pacheco Pass would be okay to add if Monterey Bay were fully well developed all the way to Salinas and included Watsonville. (Which of the two cities to connect with a new route to Gilroy and Pacheco Pass would be the big question.)
The appeal of the high-speed rail system in the Central Valley at the smaller cities that really don't merit high-speed service, but would merit other service on a real, full state railroad system, as has been said before, would be to get to and from the larger cities, and as an alternative to air travel, especially in winter. It would be met by trains going nearly as fast, but less costly, at lower basic fares.
Is it viable to extend the Gilroy trains north to Santa Clara Great America? A lot more jobs in the area than Diridon
ReplyDeleteOperationally/logistically? No problem. The trains are just ... idling. [OK you might want to add one switch and a few hundred feet of track to store your idle train while it lethargically wonders what it will do for the the next seven hours or so. Maybe just ... depreciate? Oxidate? Decisions, decisions. But there's all the time in the world to decide!]
DeleteTechnically? No problem at all.
UPRR-istically? UNPOSSIBLE. Tens of millions of payoff before we'll deign to consider your ridiculous proposal. And after pocketing the tens of millions? Nope. Can't happen. Un-possible, after all, it turns out. But, hey, feel free to try again!
San José, Capitol of Silicon Valley, delusional-y? NO F*CKING WAY.
There's a siding already in planning at Great America specifically for the purpose of stabling trains. Not for Caltrain specifically, but there's no reason it could not be used that way.
DeleteEven better, why not have one of the Gilroy runs be taken over by Capitol Corridor and run it all the way to Sacramento?
DeleteOf course the problem is that, since CC/Caltrain are different agencies, it is a physical impossibility for them to coordinate a timed, guaranteed, cross-platform connection.
I might misremember the details, but in previous discussions in the comments here (for this or some of the last few posts) someone posted a link about a study to improve capacity for passenger trains between San Jose and Oakland, in order to let the Capitol Corridor always run all the way to San Jose rather than for the most part terminate at Oakland. If the suggestion in that study gets built, you could either run Giroy trains all the way to Oakland or even let them become part of the Capitol Corridor.
DeleteBut: Isn't the idea that the few Gilroy trains run all the way into SF?
Yes - the Capitol Corridor vision plan is the one you are referring to MiaM - link: https://www.capitolcorridor.org/vision-plan/.
DeleteRE: AG's comment - that is a great point. I know that the Salinas extension for passenger rail is underway (https://en.wikipedia.org/wiki/Monterey_County_Rail_Extension). I'm not sure if they're partnering with Caltrain or with CC, but there's a lot of overlap there. I know that this extension ideally wants a couple of trains that go all the way from Salinas to San Francisco, but I'm not sure how realistic that is, especially with CAHSR and electric wire.
My ideal solution would be some sort of partnership between Capitol Corridor and Caltrain once Capitol Corridor moves to the Coast Subdivision. They could do some sort of regular service between Salinas, Gilroy, San Jose, Santa Clara, Oakland, Sacramento, with cross platform transfers at Diridon and Santa Clara for San Francisco bound riders. They also could turn some trains back at Diridon and some trains back at Oakland, also. This would be a great way to slowly upgrade the Coast Subdivision between San Jose and Gilroy in anticipation for CAHSR, since the trackage is very straight and would just need grade separations and 3-4 tracks throughout for 125mph diesel service (if UP allows it). From there, stringing up wires wouldn't be that heavy of a lift!
Now that I think about it, AG has an excellent point.
DeleteIf the negotiations with UP for the Tamien - Gilroy segment go well, plus the other upgrades (like Caltrain + Capitol Corridor; Monterey extension, etc) start to come down the pike in advance of CAHSR, that would be very interesting.
A potential plan for the Tamien - Gilroy corridor is this:
1. Close certain grade crossings - I count at least 4 or 5 that could be closed today with minimal to zero impacts.
2. Upgrade the existing crossings and some trackage to 110mph standards, including some double-tracking, passing tracks, and quad gates. This wouldn't be too expensive, because no electrification will happen yet, and the trackage is very straight as is. Again - only about 16-18 crossings would require quad gates and 110mph equipment, making this relatively cheap.
3. Slowly grade separate, in concert with upgrades for stations. I'd prioritize the sections in between stations, as that is when trains would be at their highest speeds, and the grade crossings in town are usually around the stations, where the trains are already going very slow. I am thinking a split between the feds, state, CAHSR, UP, and cities where there's a budget of $50M to $100M per year until this is done which is very doable. The grade separations should include some sort of barrier wall or intrusion barrier to separate the freight tracks from the passenger tracks.
4. In parallel with grade separations, future-proof the upgrades to properly site foundations for poles and wires, plus acquire land or adjust for electrical substations. Also, future proof for 186-220mph operation when and where possible (passing tracks through stations, curve straightening) which should be cheap as the route is largely through farmland and is very straight.
5. When grade separations are nearly done or done, start running at 125mph, using existing diesel like the Siemens Chargers. This reduces the runtime from Gilroy to Diridon significantly, giving us about 35-40 minutes at 125mph. Caltrain, Capitol Corridor, the Monterey extension, and even Amtrak would benefit from this significantly.
6. By then, CAHSR should be working on the Pacheco tunnels, and should fund the rest of the work to electrify the route from Tamien - Gilroy.
7. If the line is already future-proofed for 186-220mph operation, this cuts off as much as 20 minutes off the runtime for CAHSR, which is not peanuts. I really think CAHSR should be looking at these sorts of iterative upgrades to improve overall speeds, reliability, and service.
Costs:
Assuming 18 grade crossings, following an average cost of $200M from the peninsula is $3.6B, but that's a very high estimate. Construction costs would be lower along this line, with less impacts because the towns and cities are much smaller, and there are some rural grade crossings which are easier to fix than on the peninsula. I'd say it's closer to $2B, and with a $100M budget per year, that gives us about 20 years to get all of this done. Plus, if it's a broader strategy, we can see a standardized approach with standard designs and like in the Central Valley, standardized beams and berms that can just be copied and pasted for each separation. If more funding comes down earlier along the line (e.g. federal grants), then we could get this done in just 5 or 10 years, but $100M a year between Santa Clara County, Gilroy, Morgan Hill, San Martin, UP, Caltrain, CAHSR, and maybe Capitol Corridor is very doable, and the benefits would be huge, especially for CAHSR in the long run.
A question is if the Captiol Corridor should be ran by Amtrak or by some "Norcalitrain"? What are the pros and cons?
DeleteIf decisions re rail could be taken at state level, a trump card that the state could have when negotiating with freight railways is allowing exemptions from stricter pollution requirements for their diesel locos. A decent diesel loco that is a few decades old is likely still better for the local environment, and definitively better for the global environment, than if the freight would run on tracks, and thus it's reasonable to allow exemptions. But more importantly, if such exemptions ends up with allowing more passenger trains on the freight company owned tracks, then bringing people from cars to those trains will most likely reduce way more pollution than the harm from exempting older diesel locos.
Btw, a requrement though should be a limit of for how long a loco are allowed to idle. I get that some manufacturers don't even allow more than 1-2 cold starts within the warranty period of locos, but in that case just use a heater to keep the engine/oil warm while the engine is off.
Don't know to what extent locos are left idling within California though, but still.
The website for "Norcalitrain" is https://www.capitolcorridor.org/ "Capitol Corridor Joint Powers Authority (CCJPA) is a partnership among the six local transit agencies in the eight county service area which shares the administration and management of the Capitol Corridor. The San Francisco Bay Area Rapid Transit District (BART) provides day-to-day management support to the CCJPA." ...and I believe Amtrak crews run the trains.
DeleteIt has to be repeated with the following:
DeleteUPRR-istically? UNPOSSIBLE.
Ho, ho.
The Capitol Corridor Web site indeed is the place to go, including for future plans, or better said, hopes or dreams, not even considering any direct SF-Oakland connection and related SF-Sacramento service one day. Franklin Canyon was thankfully included as a potential improvement in the East Bay. Interestingly, there's speculation on the wild side of revival of the Sacramento Northern passage of Solano County as a new freight railroad(!), but it's also amusing for passengers given it goes through an area of recent special notoriety.
Deletehttps://eastsolanoplan.com/
There is talk already of extension to Salinas in some form, and the whole area around Gilroy in particular, to the eastern flatlands beyond 152 & 156, often-neglected Hollister, San Juan Bautista, back around to Gilroy is likely to be developed, and has room for building the housing that's wanted, more houses. (It's narrower toward San Jose, and north of Morgan Hill there is land preservation.) There's no reason the area's future need become all New-Cal metro periphery warehouses and another truck stop "travel center," maybe a casino or card room.
Great that the Capitol Corridor is at least a local transit joint effort rather than something that Amtrak runs fully on their own.
DeleteRe the UP letter to Cali HSR:
If what they say, that land owned by UP in Gilroy is set to be used for industrial/commercial, wouldn't HSR be able to use eminent domain on that part?
That letter also showcases why railway lines, meaning the infrastructure, shouldn't be owned by private companies.
But also why doesn't FRA do more when it comes to freight derailments? Sure, those happens all around the world, but to me it seems like freight railroads in USA have a "brutal" approach in that they put less effort in preventing them and treat them like something that happens in USA, kind of like that North America has tornadoes while Europe more or less don't.
I put together my thoughts on this discussion here on Reddit: https://www.reddit.com/r/transit/comments/1fg4d9h/postelectric_service_for_caltrain_merging_capitol/
DeleteMy idea is that Capitol Corridor should take over the Caltrain service south of San Jose and integrate it into their line(s). This is partially in anticipation of the CC extension to Salinas, but also to provide a regional and fast connection from Salinas to Gilroy to San Jose (and Santa Clara/Great America) to Oakland to Sacramento.
@MiaM: Eminent domain would result in a bigger battle and more cost, including for compensation for lost prospective (exaggerated to imaginary) future business for UPRR. Sadly (as you know with Amtrak disruption by freight trains), while cars are private and run on public infrastructure with public traffic control, and air travel has that model, there is no "parity" enjoyed by rail travel with this. The tracks are almost all private and dispatching is private. And private railroads have their own interests that don't coincide with public passenger rail projects.
DeleteThere has been a recent change in railroad operations that began with one person with a name for what he introduced (Precision Scheduled Railroading) with other railroads following in turn, in a time of shareholder power and greed and the same and catering to it by executives who get rewarded themselves for cost cutting of all kinds under what has become a huge umbrella of a new term for pleasing the stockholders in large part from cost cutting, including with maintenance, as critics are happy to tell everyone else.
“My idea is that Capitol Corridor should take over the Caltrain service south of San Jose and integrate it into their line(s).”
DeleteYes, exactly this. The distance, urban form, and population of the corridor from Morgan Hill to Gilroy (and on to Salinas, Hollister, etc.) is suitable for the regional rail service provided by Capitol Corridor, not the commuter rail service provided by Caltrain.
@Anonymous:
DeleteIIRC I've brought this up previously, but can't remember what the result of the discussion was, if there even were a specific discussion.
Is it technically possible for a US state to tax a specific type of business? If so California should consider taxing the freight railway companies so they either sells their infrastructure to the state or go bankrupt. Taxing publicly owned railways isn't a problem as the tax would just loop back as subsidies for running public transit.
Or probably better put a hefty tax on owning major transport infrastructure without not only allowing but also prioritizing local and regional passenger transportation with a reasonable fee for track usage.
MiaM: If the intention is to take over the railroad, adequate compensation remains full compensation; nobody competent is going to be fooled by any passive-aggressive substitute. Then there are other questions like interference with or discriminating against interstate commerce, for example.
DeleteEven taking over dispatching to ensure passenger priority may not work if it's at the expense or to the detriment of private business.
@MiaM: You were wondering aloud about an idling limit. Here you are, for California; it's lost within new zero emissions locomotive requirements (the New Edict) insofar as reaction and news are concerned.
ReplyDeletehttps://ww2.arb.ca.gov/our-work/programs/reducing-rail-emissions-california/locomotive-fact-sheets
https://www.trains.com/trn/news-reviews/news-wire/california-concludes-zero-emissions-locomotives-are-ready-to-roll/
More on idling reduction can be found here:
https://www.epa.gov/verified-diesel-tech/learn-about-idling-reduction-locomotives
If you visit the Capitol Corridor site, the dream is electrifying the route and the best rail routes being electrified are a future possibility, but it's premature now. With all passenger rail, the best routes need development and completion first! (Including the missing parts, such as the new Altamont Pass, Dumbarton added, a new Mission Pass route to connect the Tri-Valley and Fremont and Dumbarton, crossing the Bay to connect San Francisco and Oakland...)
That's obviously different than the real application of cleaning the air with the New Edict, to clean freight movement.
With people's motor vehicles, new models can feature stop-restart to save on fuel consumption, and in theory should reduce emissions, too. (Efficiency and emissions, like smaller displacement with turbocharging that was overdone) Stop-restart is rejected by a good number of people, though, for their vehicles. (Clumsy or threatens premature wear)
As for NOT driving like a grandma, why can't there be a full-power, full-speed demonstration of Caltrain for us?
ReplyDeleteIs there concern things won't work as advertised and particularly that there might be breakdowns or electrical problems?
After all, such "driving" should have been fully tested and wrung through, even at excessive power levels and maybe even excessive speeds (not to mention acceleration) to demonstrate ability and competence and [gasp] reserve, before fielding the trains.
All the max power draw and contingency scenario testing was done during the whole weekend shutdown back in June. A week and a half is just another blink to wait at this point
DeleteSiemens Mobility to establish America’s first high-speed rail production facility in Horseheads, New York
ReplyDeleteSo much for Siemens’ possible “manufactured in California” advantage on the CA HSRA’s trainset procurement.
Yep, so much for that. Schumer gets to gloat now; it's added to Alstom.
Delete(Upstate, Southern Tier)
https://www.stargazette.com/story/news/local/2024/06/12/brightline-west-train-manufacturing-should-set-up-shop-in-ny-schumer/74048861007/
https://www.the-leader.com/videos/news/2024/09/09/schumer-talks-impact-of-new-siemens-rail-manufacturing-in-horseheads/75147326007/
I still wonder when eastern and western railroads might merge, "for efficiency" now.
ReplyDeleteWould the combinations be UP and NS, BNSF and CSX?
Video of an EMU pantograph failure illustrates how a train can continue using the other pantograph, but raises the question of why a pan was destroyed on brand new infrastructure and whether a root cause has been identified. Caltrain's famous maintenance department is surely on the case using their rigorous failure analysis processes to prevent recurrence! Failing that, October is gonna be... interesting.
ReplyDeleteThat's ... Not Good. Not Good At All.
DeleteOof.
Assuming the OCS and EMUs are still under warranty, it is the responsibility for Stadler to work with Balfour Betty to investigate why the pantograph failed. Of course, for now at least, they'll replace the pantograph and marked this under isolated incident, would only investigate further if there are more occurrences.
DeleteWilliam, have you considered lobbying your local county to get yourself appointed to the Caltrain JPB Board?
DeleteYou have the sorts of insights that would fit right in, helping Caltrain be Caltrain.
At what point (approximately) in the video is the pantograph failure seen?
Delete(The video is almost 20 minutes long and it seems like the description doesn't point this out).
Btw given how cheap surveillance camera systems are nowadays it's a good idea to have cameras that record what happens to the pantograph.
I assume that the trains, like I assume any modern vehicle, have auto drop. That is usually done by having a cavity in the carbon thingie on the pantograph that contains pressurized air, and if the pressure drops the pantograph is automatically pulled down in order for damages to not get worse.
There is a classic video online of pantograph failures with and without auto drop. The one without autodrop continues to "work" somewhat, i.e. it conducts electricity, while simultaneously both damages the infrastructure and gets more damaged during a minute or two until it fails more spectacularly. The one with autodrop just drops down and that's it. Sure, in both cases the wires need to be inspected as there is a decent chance that the damage is caused by some infrastructure failure, but it's way easier to repair a minor damage than lots of damage spread out over a longer section. If the exact time of the auto drop is correlated with exact positioning it might even be possible to order trains to coast through that area with their pantographs dropped until there is time to inspect the wires.