Photo by “Beast Mode” AVGeekJoe / flickr
Link Light Rail had its busiest year by far in 2016, and saw its largest ridership growth ever, with ridership growth surpassing the original year’s ridership for similar periods of time, counting from the opening of University of Washington Station and Capitol Hill Station. It is still settling into its new routine after the opening of Angle Lake Station in late September.
Calls continue for running 3-car trains as much as possible, to deal with occasional but frequent crushloads. Having only three-car trains could probably be done, at the same scheduled headway, at great expense, and possibly messing up some plans for what happens after the buses get kicked out of the Downtown Seattle Transit Tunnel in the next few years (as, once ST rolls out service, frequency, or capacity, it is loathe to undo it later).
Today, I’ll offer some suggestions to improve Link operations that can be done regardless of the 2- or 3-car train debate. The first four impact what will be in the printed schedule that comes out in March.
1. Remove the minute of padding in the schedule at SeaTac Airport Station. It does not take four minutes to get there from Angle Lake Station or from Tukwila International Boulevard Station. We have the video evidence that ALS to SAS can be done in 2:45. Nor does it always take at least a minute for passengers with luggage to board or deboard. Thanks to the schedule padding at SAS, operators often have to wait half a minute or more after everyone boards before the scheduled departure time allows them to proceed. Some operators keep the doors open, and some keep the doors shut, while waiting for the departure time. If drivers could save half a minute each way at SAS, that would be another minute of break they could get.
2. Differentiate between peak and off-peak travel times. Thanks to higher ridership and longer dwell times, peak trips take longer. The schedule ought to reflect that. Not only would it give riders more accurate information, but it would also reduce the chaos of trying to maintain peak headway when drivers are perpetually behind. This applies not just to Link, but to every bus route.
3. Publish and announce the fact that only 3-car trains are running on all weekends and holidays. If in the future ST decides not all weekends merit all 3-car trains, it can publish that change in the schedule. But few are expecting the 3-car trains, even on weekends, since it isn’t being announced, so the third cars continue to be undiscovered stretch limos.
4. Try out having only one southbound bus bay in the DSTT. Route 106 left the DSTT permanently in September. Perhaps Bay C can handle routes 101, 102, 150, and 550. I’d be surprised if it can’t, given that Bay A has 26 outbound buses and 13 inbound buses during the PM peak-of-peak hour, while Bays C and D each only have 12 outbound buses, and 9 inbound buses between them, during that time. It would be nice to test the possibility for a week sometime in the winter.
| Route | Northbound (Bay A) | Southbound (Bays C & D) |
|---|---|---|
| 41 | 14 | 4 |
| 74 | 4 | 0 |
| 101 | 3 | 4 (Bay C) |
| 102 | 0 | 4 (Bay C) |
| 150 | 4 | 4 (Bay C) |
| 255 | 8 | 5 |
| ST Express 550 | 6 | 12 (Bay D) |
| Total | 39 | 33 |
Removing Bay D would clear the platform space for passengers to wait for the second train car (currently partially blocked by the 550 riders queue), and reduce crushloads on the front car, where they occur most. It might also make it easier to install between-car barriers.
If Bay C cannot handle all four routes, then have routes 101, 102 and 550 at Bay C and 150 at Bay D.
5. Install floor markings on the platform that show where the doors open, and that encourage boarding riders to stand aside while other passengers deboard. Do it for where the doors would be for 3-car trains, too, but have verbage that not all trains have three cars. (This is already in the planning, but so far just as a pilot project in Westlake Station, and only through the end of joint operations.)
6. Install last-departure count-up clocks at University of Washington Station and Angle Lake Station. These would be similar to the clocks at Mt Baker Station and Rainier Beach Station. Headway control is one of the keys to avoiding unnecessary crushloads.
7. Install between-car barriers in all the stations, not just because it is the right thing to do for passenger safety, but also because the barriers help passengers position themselves to board either the second or third train car, depending on how long of a train shows up.
8. Make fuller use of ORCA Boarding Assistants. The boarding assistants are drivers who work for King County Metro, which is responsible for DSTT operations. They can be asked to do things like let the waiting crowd know whether the next train is 2 or 3 cars long. I continue to see crushloads on the front car while the third car is quite spacious. Letting the crowd know there will be a third car would allow passengers to migrate thusly on the platform, and make better use of the third cars.
The boarding assistants can also assist in getting passengers to get out of the seats that have unfortunately been installed in the wheelchair space so that riders in wheelchairs have a much smoother boarding experience. [I’d suggest removing all the seats in the wheelchair spaces, but that is a macro fix.]
9. Have 3-car trains leave University of Washington Station up to a minute late whenever they are alternated with 2-car trains. That will help even out the loads, give in-the-know riders a signal to migrate to the rear-car waiting area on the platform, and reduce the train bunching caused by alternating 2- and 3-car trains.
10. Extend the mobile ticketing pilot project at least through the rollout of ORCA 2.0.
11. Do as the county council requests, and get rid of the nation’s most-expensive-by-far fee for acquiring a bus smart card (the ORCA card). If some sort of incentive is needed to keep riders from treating the card as disposable, it need not be more than 25 cents. $5 is EXTREME OVERKILL for the purpose, that has forced a lot of expensive workarounds for occasional and cash-flow-challenged riders.
12. Upon vanquishing the $5 ORCA CARD fee, require use of a fare medium other than cash when boarding a bus in the DSTT.
.
I’m not taking sides on whether 3-car trains should run all day. ST has Service Standards and Performance Measures that help guide that decision. But these remaining suggestions apply if ST goes that direction:
I. Give drivers longer breaks during peak at Angle Lake Station in exchange for trading off trains, so the loop can be done with as few trains as possible. Because there are only 62 light rail vehicles in the fleet until at least 2019, and 10% have to be reserved for maintenance, minimizing trainsets in the peak loop is the key to maximizing 3-car trains in the peak loop.
II. Make sure to install the clocks suggested in item 6 above so that drivers who arrive at a terminus after the previous train leaves will have some guide as to when the previous train left.
III. Run all 3-car trains always, announce and publish that all trains will be 3 cars, and deploy the floor markers, so that riders will take advantage of the third car. The difference between running mostly 3-car trains and all 3-car trains is significant in operational impact. It might take 18 trains to run the loop, even with the improved smoothness from all 3-car trains. That still leaves 8 spare LRVs, more than enough to keep the 10% spare ratio. It doesn’t allow for a spare train. So be it. The smoothness from getting riders used to 3-car trains ought to outweigh the occasional hiccup from not having a spare train. Hopefully, running the loop with 17 trains becomes feasible as DSTT throughput improves.
If ST moves to all 3-car trains off-peak, it should do it right and go all-in, not just do it halfheartedly to assuage complaining riders.
13. Get escalators that work.
I’ve missed far too many transfers due to the UW broken escalator zig-zag costing time. Most people also go into the first available car, so the one right next to the working escalators is crush loaded, while the far car is nearly empty.
This. ST has yet to give a satisfying answer for the frequent escalator/elevator breakdowns at Husky Stadium and CHS. Are they performing as expected? Is this some break-in period or is it going to be a daily surprise to see which escalators/elevators are broken forever?
Would like to hear an answer from either Sound Transit, or the elevator vendor,or both under oath in open court. Or at least an explanation, delivered by a technician rather than Customer Services, as to why exactly our elevators and escalators can’t either work or be repaired.
The one at Sea-Tac Skybridge needs an ADA-related citation. Critical location, many transferring passengers with more luggage than they can carry. How long has it been out of order this time, and how many times before?
Tell me, someone who knows: is necessary enforcement local, State, or Federa? And civil or criminal? Would rather have elevator vendor in jail or sued out of business than ST.
Because if the law destroys Sound Transit, that’ll means we’ll have out or order signs at every stop from Everett to (rush hour only) Olympia. And IDS to Redmond too,
Mark
Any normal transit agency would aim for at least 95% reliability with the escalators, same as the trains. 60%, 50%, 40% reliability is completely unacceptable. The break-in period should have been before the stations opened. Every passenger who rides a train also has to get to the platforms.
I head to the elevators if I see a stopped escalator or hlaf the time when I don’t, because you never know when you’ll encounter a stopped escalator at another level, and there’s a 50% chance you will, and I can’t tolerate walking up or down escalators or switching to the other side mid-level any more.
If you are trying to get to the platform at UW Station, take the elevator down. It is never crushloaded in that direction, and beats the escalators even when all are working. Insistence on taking the escalator down is an illogical reason not to pack the rear car. I see lots of people taking the elevator down, and it never seems to be a problem.
Of course, not everyone can take the elevator at once trying to exit the station after deboarding. That’s why I keep asking that the partner escalator get set to the egress direction when one is broken down.
Elevators can bear most of the passenger load from people showing up randomly and steadily to get to the platform. Escalators are needed (and overwhelmed) for getting away from the platform.
Before U-Link I hardly ever used the elevators to save electricity and expenses and keep them free for disabled people and bicycles. But U-Liknk is making it really hard to do that.
When we flew out of Seatac before Christmas, one of the down escalator was not working. It actually took over a minute to empty the car we were on because there was only ~5′ between the car and the side of the working escalator that everyone had to funnel towards, creating a traffic jam.
You’d think that ST would prioritize getting escalators working on days they know to be very high traffic.
If a department store had elevators that failed this often, the company or individual responsible for maintaining them would have their contract terminated with a replacement hired promptly. Can ST hire whoever maintains escalators for Macy’s, Nordstrom, and Southcenter??? If they haven’t already done so, they need to write a termination for non-performance clause into their next contract.
While I’m no expert on moving conveyances, I’d imagine the elevators/escalators at any typical department store or shopping mall don’t have quite the heavy usage as one at a public transportation facility. For example, the ones at shopping malls live in temperature controlled, indoor environments. While a ST escalator, while covered isn’t exactly indoors or temperature controlled. A shopping mall will be lighter usage, indoor, and probably turned off at night, whereas ST’s probably just run 24/7 and don’t get shut down at the end of the service day.
“occasional but frequent crushloads” – I do not think you can have both occasional and frequent.
Sure you can. “Occasional” here means “irregularly spaced,” not regularly occurring. And frequent means that the crushloads are common.
Then it should be “irregular but frequent”.
‘Occasional’ connotes whimsy, and a crushload is the very opposite of whimsy. Unpredictable is the word you are looking for.
I strongly support the idea of always using three car trains but how about just increasing the headways slightly to have one spare train? What would the headway be at peak times? I really don’t think most riders would even notice. I’ve gotten in the habit of not running to catch Link trains during peak periods because they are so frequent currently.
I’ve wondered that too. Just 60 or 90 more seconds between trains could seemingly free up enough train sets needed to run all three-car trains. There aren’t any timed transfers, so that couldn’t be the issue.
The only issues that I could see is that there is something about the MLK signal timing or the DSTT sequencing that sets the headways. Otherwise, it’s a only a fierce clinging to a political promise.
Just to add a dissenting opinion: I would notice.
I hate waiting and that awful feeling of approaching the station just as the train pulls away. Six minutes is right at the cusp of where I can mostly relax and not stress about it (3-5 minutes would be better.) An additional 90 seconds would be a 25% wait time increase during peak operations. That is substantial.
I would notice too.
“An additional 90 seconds would be a 25% wait time increase during peak operations. That is substantial.”
So the 2-3 extra minutes it takes the train to travel through the tunnel “due to traffic ahead” is not substantial? By the time your train gets loaded, 30 of those precious seconds have been wasted pushing people to make more room to let everyone on. It’s like the new PAYSTTE, except it’s during boarding, and in the place you least want trains holding up buses.
It wasn’t a political promise, just an annoucement that ST would use shorter headways rather than longer trains to deal with U-Link capacity, It could change it anytime. What ST has “promised” before 2021 is 10-minute headways off-peak, 7-8 minutes peak, everything beyond that is extra.
I should’ve also added that SDOT signaling is having a heck of a time accommodating the relatively new shorter Link headways along MLK (about the past year?). Signaling hardware was recently upgraded/replaced at some intersections but my understanding is old code is still in place that struggles to let trains, car turns and cross MLK movements happen along with pedestrians crossing MLK. Increasing headways slightly could only help this issue too. SDOT works with ST to give Link priority along MLK but it’s still a work in a progress as just this morning I witnessed warning signals activate at Edmunds after a northbound train had left the Columbia City station (no train at Edmunds in either direction).
I’m with you on that one, Scott. If ST were to roll back the scheduled peak headway to 7.5 minutes, in exchange for 3-car trains all the time, I’m pretty sure those 3-car trains would be ample capacity on most days, including M’s days and Sounder weekday matches. (This is in part due to lots of sportsball fans travelling in the counter-peak direction.)
People would have more comfortable rides. Sure, they’d have to wait up to 45 seconds more on average (less than that, probably, as headway control improves, so it might even be a wash), and then they’d have a faster ride, as dwell time would be cut by up to 33%, there would be fewer delays along MLK, and the DSTT should be pretty much free-flowing with the reduced stress. I’m pretty sure riders wouldn’t mind waiting a pretend-extra 45 seconds for a real shorter trip of 2-3 minutes with real more spaciousness.
Riders on all the tunnel buses would notice a substantial improvement, too. Even if ST doesn’t care about Metro’s budget, some of that savings would accrue to the ST Express budget, which is stretched to the max. Imagine being able to run an extra 550! It would be full at rollout.
Running 3-car trains all the time, with 7.5-minute peak headway, would certainly cost a lot less in maintenance and mileage than doing the same with 6-minute headway. Part of that would be due to keeping a full maintenance staff busy during peak. Plus, the drivers would have an easier time getting to stick with their train.
My napkin math says 15 cars could fulfill the peak loop at 7.5-minute headway, compared to the current 18 or 19 trains being used.
ST simply wanted to avoid the maintenance and mileage costs of running 3-car trains all day, half empty. They gambled that 20 cars per hour per direction during peak would be sufficient, and lost that gamble. So, they run 23.5ish cars per hour now, with 10 of them near-crushloaded, 10 SRO, and 3.5ish mostly empty.
Thanks! That’s the analysis I was looking for. Basically going back in time a year in terms of headways but with three car trains. All the points you mentioned are important. Several constituencies benefit.
Number 11: Orca card high cost is certainly absurd. The problem is likely that someone relies on that money. Has anyone followed the money? Is there an elected advocate who will step up to the plate and push to lower or abolish this $5 cost? Given the car tabs, property taxes and sales taxes we just promised to give ST, a free or reduced Orca card shouldn’t be difficult to provide.
The income from selling the cards helps cover the cost of the vendor contract. They don’t seem to have found an easy way to apply money saved from efficiencies to pay off the contract instead. The vendor charges somewhere between $2-$3 for each card, and then ST/Metro make a $2-$3 theoretical profit. The devil is in the detail that they are losing a lot more money than that for every semi-frequent rider who opts to keep fumbling cash.
The King County Council budgeted $1 million over the next two years to buy out the card fee. But first, they funded a study, and then the ORCA pod has to be convinced.
What’s process called that’s the reverse of “Privatize?” Somebody also tell me how paying shareholders’ profits in addition to operating costs is less expensive than a public agency just doing the job itself.
That kind of accounting doesn’t sound very conservative to me. Somewhere close upwind, a wrong ideology has fallen into a broken piece of machinery and been rotting there for about forty years.
Mark
As I understand it, ORCA 2.0 will be administered in-house by the ORCA pod.
“3. Publish and announce the fact that only 3-car trains are running on all weekends and holidays. If in the future ST decides not all weekends merit all 3-car trains, it can publish that change in the schedule. But few are expecting the 3-car trains, even on weekends, since it isn’t being announced, so the third cars continue to be undiscovered stretch limos.”
What about parking 3-car trains half a car length farther up the platform than 2-car trains? So people don’t have to guess about whether it will be a 2-car or 3-car? If you guess wrong, you wind up piling onto a packed 2nd car. I get that it would be somewhat more complex than their current setup, but dang…
I’m not sure but I believe ADA regulations require them to stop in a consistent spot so that blind and mobility-impaired folks can anticipate where they need to be without scrambling at the last second. At least I remember something like that in regards to which bays the buses use.
Regardless of the regulations, ST is not going to budge on changing the front position of the trains until 4-car trains get rolled out, which isn’t planned until Northgate Station opens. The doors have to line up with the tactile strips, and having vision-challenged passengers step off the platform right between cars (as has happened several times around the US&A) is a serious risk.
ST is getting ready to install between-car barriers in the non-DSTT stations, which will cement the need to pull the train to the same stop position within a foot or less of imprecision tolerance.
Hopefully, those barriers will be as long as possible, given that the narrow stop range is causing some drivers to pull into stations reeeeeaaaaaallly sloooooowwwwwwllllllllllllllllyyyyy. I was on a train where the driver seemed to take almost a minute to complete deceleration at each station, and ended up taking 54 minutes to do the whole trip. Off-peak.
Brent –
That’s a training issue and should be reported to be addressed.
The New York City subway has to stop all of their trains in a precise manner. Trains that are much longer, much heavier, and approach the stations at much higher speed.
The motorman must stop the train such that the conductor (in the middle of the train somewhere, depending on equipment type) has his window lined up with a small, zebra striped board on the ceiling.
This level of precision is at least, if not greater than, what the LINK operators will be required to do with the platform barriers.
It should not be a big deal.
If you eliminated cash fares on buses inside the DSTT, I am pretty sure someone with x disadvantaged rights group is going to complain that their represented clients will face x undue hardship because of ORCA’s X unfair advantage.
If someone still finds putting $5 onto an ORCA card to be a financial hardship, they probably qualify for the ORCA LIFT card, which is free. Every bus and train in the tunnel has a LIFT fare of $1.50.
Numerous human service agencies give out combo Metro bus tickets / Link day passes to people who truly can’t afford to pay any fare.
If someone has a disability, they probably qualify for the Regional Reduced Fare Permit, which would become free if the county council gets its way. All trains and buses in the tunnel have a $1 RRFP fare.
Did you have any particular edge cases in mind?
I don’t think it’s much of an edge case problem. The Orca card costs about as much as a round trip into the city on the train. That effectively doubles the cost of the ride. If we want people to have Orca cards they should be cheaper.
Though at this point I’d rather see a focus on allowing the use of debit or credit cards directly and NFC or app-based payments.
How about the general case of having to pay accounting and administration fees over and above actual cost? What are we, as a transit system and its passengers and voters:
Cheap, lazy, scared, or stupid? Or just used to getting robbed to pay somebody’s profit for things we the public can do for ourselves? Prosecution rests.
Mark
The goddamn paperwork for “special cards for the poor” is its own expense, costing hours of time when you COULD be working and earning money.
Just get rid of the damn fee. What the fee does is to chase away occasional riders, meaning only regular commuters will take the train. That’s stupid.
If you got rid of the fee people will just toss them out after a single use, and it will cost even more money.
Eliminating cash fares in the tunnel doesn’t necissary mean 100% ORCA use. It could mean ticket printers for Metro as well as SoundTransit.
If the Ticket Vending Machines are currently ADA compliant there’s no reason they can’t spit out a ticket for Metro or ST Express Buses just like they do for Link. Show up at the tunnel without ORCA and you are just forced to use the TVM. This would have the added benift of allowing the tunnel to transition the platforms to a fare paid area.
Using the ST TVM’s to print out dual use tickets is probably doable, albeit somewhat expensive to reprogram. Probably in the tens of thousands range. Ticket stock may be an issue as well, since last I saw (and its been a few years now) the ST tickets have the wave pattern pre-printed on the stock, which means you would either have to load two stocks, or use a more generic ticket stock for the DSTT. Then you would have the hassle of training the riders which is another story all together.
The ticket design would probably need upgrading. They are designed for FOE’s to take ahold of and inspect. Having operators accept them as flash passes on buses wasn’t the intended use, will be fraud-prone just like the paper transfers, and would escalate the risk of disputes on buses that don’t have time for them.
The upgrade project might just be completed by the time the buses get kicked out, so I don’t think it is happening.
In the grand scheme of things, I’d rather see ST put its long-term effort into developing a plan to transition to open gangways, and figuring out how to maximize the length and capacity of said super-trams.
From the article posted in the STB not long ago about open gangways, the estimated overall capacity increase was not that great, something like 40 people which makes it difficult to justify the additional costs of a custom design, and modifications needed for the shop facilities to handle single trainsets of that size. And we wonder why things cost so much to build…
The capacity might only be 40 more (I suspect that is an underestimation, as ST seems intent on double-cab cars), but the number of people who will practically fit is a larger difference.
If we had 300-foot open gangways instead of 3-car trains, we wouldn’t be talking about frequent crushloads.
Nor is there a reason the open gangways would be limited to 400 feet, as they could extend beyond the platforms, so long as the doors are within the platforms.
Add in the savings from people just boarding instead of running for other doors when they can’t get on the car they are trying to enter, and you have dwell time savings. Reduced dwell time means reduced minimum headway. Reduced minimum headway means increased throughput. Increased throughput means increased capacity.
The big savings is not having to build a third downtown light rail tunnel, or at least pushing it back decades from if we just stuck with old-school trains, assuming the engineers are bright enough to build three-platform or four-platform stations in the second tunnel.
I also still think it is not too late to retrofit the DSTT to have three platforms per station. Once the cutover to the second tunnel happens, that would be the ideal time to remodel the DSTT stations.
Here’s an issue with clocks: Figuring out where to install them. The trains, after all, do utilize both sides of the track.
Can’t you just stick one on each side?
As car as I’m aware, the headway clocks reset as a train passes by the device. Theoretically, we can have clocks on both sides of the track. But then it would super corny to have the things reset every 20 minutes a train leaves from one side. Or…however long it takes.
Set the clocks above the end of the platform. Have two sensors that can trip the clock. Test it for a week or so.
If necessary, have a CPU feed two clocks working off the same data.
Like this hasn’t been solved in some other city of the world. Why do we think we’re inventing this stuff for the first time?
A printed Link schedule is coming out in March? Seriously? With trains as frequent as they are, printed schedules (timetables) are simply unnecessary and will be rarely used. And even the slightest service interruption (which happen almost daily) will render them useless. Sound Transit should publish the times of the first and last trains each day, and the service frequencies. That’s all that’s necessary and useful.
What’s wrong with LINK schedules already on station information boards and in ST schedule books? At six minute headways, passenger won’t even have found the page when train arrives.
Mark
I would like a print schedule because even yesterday, I had to yank one out for the Everett Transit bus and recalculate due to an unforeseen delay how to make the multi-transfer trip back to Skagit.
But once the frequency gets down to 6 minutes or less from 6 AM until 10 PM, sure.
We went throught that before with both Link and RapidRide. People want schedule-based operations, not headway-based. That way you can figure out which train to take to transfer to a 30-minute bus, or how long you’ll wait. Having trains show up at at random times is difficult if you have a bus at both ends of the trip.
If headways are below 10 minutes, people want headway-based operations.
At 6 minute headways I can live without the schedules, but at 10 minutes they are both handyn and simple. (On the 2s, on the 4s, etc.)
And in my particular situation I do a lot of 1 stop trips, chs to Westlake. For these, trip time is wait time + 7 minutes (2 min each enter and exit, 3 min travel). If I can time it right with 1 minute cushion, that’s an 8 minute trip, which is a lot better than 17!
Brent, I can’t figure how how those two woolly mammoths keep chasing the stray cats and pigeons out of the Tunnel, but ’til we find the shovel shed, let’s at least put in to Animal Control.
Cute little elephant, signals on MLK. Its big 27 year old brother, King County Metro Transit’s absolute refusal to take direct control of any passenger vehicle between the (supposed to be) control towers at CPS and IDS.
Only plausible excuse: signals were all powered by a turbine in the garbage bin intended to be a waterfall fountain behind the northbound platform at Convention Place. Clocks too. Evidence: both got and were left permanently broken several weeks into operations.
Cost out twenty-seven years’ per-minute operating delays and EastLINK terminal could be New York City, Central the Florida Keys, South Tierra Del Fuego, and North, the ferry terminal for the North Pole. Re: climate change…Northgate?
Right now- key words just left of the dash. ‘Til we find the legendary fortune in wasted dispatch hardware- same closet as elephant shovels, supervisors with flashlights and radios at every signal point. Fareboxes sold for scrap steel, and put loaders to wheelchair assist.
But above all, if DSTT ops and everybody in it can’t become their own division, at leastlearn to play one on TV. If they can’t , make Tunnel bus division SDOT’s problem stat. Because no way Veolia or anyone else’s private shareholders will touch either LINK or its subway.
BTW, to stay [Not Off Topic] ’til joint ops can end with less flying rubble than current Presidency, the above control measures will make safety barriers a snap.
Mark Dublin
Brent,
Sorry I’m so late. But thank you for running my photo… especially with the tagline.
Joe