Oran Viriyincy/Flickr

In the last open thread, Brent critiques Downtown Seattle Transit Tunnel (DSTT) bus operations and offers some constructive suggestions (with errata):

Observations this afternoon and evening:

1. There are some A-Line buses being used on the C/D Line. The C/D Line, FWIW, is much less full than the 120 off peak.

2. The ORCA loading assistants are getting better at using the hand-held readers, as are the passengers. The loading assistant at the southbound platform at ID Station had the drill down really well, tapping at a rate of about one card per second. I’d suggest filming him for training purposes.

3. Platooning still has some kinks to be worked out southbound and is totally not happening northbound.

4. Headway control is not happening with tunnel routes. I saw two 106s go through together, two 255s go through together, two 550s go through together, two 41s together (which is hard to avoid), and three 41s a little bit later all on the same platform at once. These were all in their outbound direction. I hope it would be easy to implement a rule that only one bus of a particular route can be deployed per platoon.

5. One of the big slow-downs was drivers opening doors for runners. I watched as a supervisor reprimanded one of the drivers for doing that. One of the 41s at the head of the block of three waited for a runner, and then his bus stalled for a couple of minutes. There ought to be a “No running on the platform” rule, in addition to clear instructions to operators to only open the doors once per platform.

6. A couple inbound buses kept their doors closed while waiting along the platform, and then opened them once they got to the front, turning themselves from the back of one platoon into the front of the next platoon, and blocking Bay A buses, which leads me to …

7. The big source of slow-downs northbound in the tunnel is that Bay A is overwhelmed. Even if platooning were happening right, passengers were boarding quickly, inbound drivers were only unloading once per platform, and the hand-held readers were functioning as quickly as the ones on the buses, there would still be delays caused by too many buses using Bay A. That list includes the 41, 71, 72, 73, 74, 76, 77, and 316. Given the rule that only two buses can be considered to be at a bay at once, and any bus further out must stop again at the bay, Bay A cannot handle the peak volume — 29 buses in one hour — unless they are spread perfectly evenly among the sixteen platoons that should be going through in that hour. And it would require all inbound buses going northbound to queue up behind a pair of Bay A buses and use the space at the south end of the platform, which is not happening.

A better distribution of the northbound peak load would be 41/77/316 (19 peak buses per hour) at Bay A, and 71-74, 76, and 255 at Bay B (18 peak buses per hour). This would require some training, rider alert signage, and a little duct tape to cover the old numbers. But I think it is a fix that can be executed quickly.

49 Replies to “Comment of the Day: Tunnel Operations”

  1. #5 is a real tough one for buses, due to drivers being human. Trains just close the doors after a fixed amount of time, but that driver has to be pretty heartless to slam the door in the face of an old woman leading toddlers. Which itself isn’t a problem, unless there’s 1-legged vetran on crutches making his way after that, etc. Ok, or regular runners, but my point is there’s a line that every human driver is going to draw and for most it won’t be to shut the door on time no matter what.

      1. Cool. How do they work? Operator watching screens? Why do they seem to always stay open well past the time everyone’s boarded?

      2. I have seen a link train wait for runners once in the DSTT northbound at university. The operator saw them running since they were approaching from the front

      3. Link has digital side mirrors, with rear facing cameras and a screen on either side of the windshield for the driver.

    1. Training just put this out a day or two ago. Word will get out, especially if they start writing PRs.

      All my joking about “being ruthless” aside, you really do feel like a dick when somebody is looking at you through your door and you pull away – plus you need to watch them as you pull away to make sure they don’t lean out to bang on your bus and lose their balance. It’s tough to get used to but I know there is another 550 only minutes behind me – plus a train, and more buses – gotta go.

      1. wouldn’t be a problem if we had next bus/train signs indicating what and when the next buses to arrive are … you know the kind of thing they should have installed when they retrofitted the tunnel …

  2. Can’t remember all the buses that used to be there, but at least post-remodel, Bay A has always been North and bay B was East. So you’d have to change the upstairs signage as well. I don’t think there’s much advantage to having them split based on which freeway the bus is using; I can’t imagine a common use case is “I need to go East so I’ll find the bay first and the route second.”

    1. Yeah this is how it has been but I don’t think many people actually know this distinction which makes it rather pointless.

    2. Again, cover up the “Bay A” and “Bay B” parts of the overhead sign, like when Link wasn’t running yet. And on the platforms, put big signs next to the bus stops saying that certain routes have moved to the other bay.

      1. I meant the mezzanine signs, but the platform signs would have to be changed too. If somebody is a visitor they’ll read the mezzanine signs; if they’re a regular rider they’ll just glance at “North” or “East (520)”. So regular riders won’t notice until they get to their old bay, which is why they need large freestanding signs next to the bus stops.

  3. The tunnel just can’t handle the peak flow and it is affecting LR operations. It is time for Metro to put their big boy pants on and start moving some peak only routes to the surface.

    1. Southbound tunnel operations have done just fine in all my tunnel travel time trials. Northbound, not so much. The only major difference between the two is the bay distribution and perhaps the tunnel entryway controllers.

      Slow northbound trains in PM peak affect a lot fewer train passengers than if southbound had the same problems, so this really is primarily a travel time issue for bus riders and an operational cost problem.

      1. I’m usually traveling northbound in the tunnel during PM peak, so I don’t want those impacts to be downplayed. It’s in everyone’s interest for the train to move smoothly and quickly through the tunnel.

  4. Some A-Line buses are used on the C and D Lines? Is that why the southbound A-Line trips are sometimes cut down to 20 minute headways during the day?

    1. If Metro needs additional buses for C/D Line they would need them in the peak, not off peak. So as long as Metro is still running 10-minute headways on the A Line during peak this doesn’t sound like the cause… unless they are extremely short on buses and are somehow doing all of the needed maintenance on RR buses during the day but putting them all out for service during the peak?

      Are you sure the bus wasn’t just late?

    2. I think this is just the usual pattern of South Base keeping the newest equipment for itself and sending the older stuff to the other bases.

    3. about south based having newer equipment that’s generally true. However with rapidride they still have the 2010 models whereas b c and d have 2011 and or 2012 models. And the newer buses will filter out to the other buses as time goes on.

      1. A lot of the 2012 LFRs are still at South Base. I wouldn’t be surprised to see them send the 2010 LFAs currently used on RR A to Atlantic and keep the LFRs for themselves.

        It’s an absolutely consistent pattern over the last 15 or so years: South Base keeps the most recently delivered equipment, East Base gets the next newest stuff, Ryerson and Bellevue get their own fleets that don’t really change, and Central/Atlantic and North always have the remnants.

      2. They will keep some of them for the f line. Others will be sent to central for use on the e line. The 2010 model buses will stay on the a line except when needed on other rr lines

    4. The presence of A-Line buses on the C/D Line doesn’t necessarily mean there aren’t also buses that were originally assigned to the C/D Line that are now assigned to the A Line. One of the things the western West Seattleites have been complaining publicly about is the lack of comfy seats. That was one problem Metro could partially solve.

    5. This was just a (recent) pattern I was noticing on the SB A-line. During the day, it would tend to not follow the proper headways. One day, when I was called home suddenly, I got out to the S 240 st stop around noon, and there was a lot of people there (a sign that it’s been a while since the last arrival), and the sign said 19 minutes. The total wait turned out to be about 23 minutes, and there was a good chance that the headway was a full 1/2 hour. It makes me want to check one bus away before going out to the stop.

  5. Operators have been instructed to not wait for runners, but it seems some operators are still doing so. Also, inbound coaches are to unload where ever on the platform and a second stop is not required at the head of the platform. Only coaches required to make second stops are outbound coaches that did not load from the first or second position in the bay. If all operators would just follows those simple procedures, that can make a big difference.

  6. This morning I saw a inbound driver from the I-5 express lanes stop mid-platform to let passengers off. Usually this doesn’t matter but and usually drivers are good about pulling all the way up to the elevators but because this bus was the first in a platoon of three the last bus (a 41) has to wait for both buses to unload before it could unload. Besides that I haven’t been in the tunnel yet because I’ve been on vacation. I think I’m going to go watch Westlake either today or tomorrow.

  7. I, for one, would not like a “No running on the platform” rule. Although if you moved 71-74, 76, and 255 to Bay B then running would be cut down as route 41 riders wouldn’t be able to run through the Bay B crowd at peak times.

    One of the biggest frustrations, at least at Westlake northboard, is getting a clear view of the route numbers as they pop out of the, uh, little tunnel (what is that connector section technically called?). If you miss the second or so straight view of it, and if there are a number of coaches pulling up, then it’s common to see people dashing all around the platform just to confirm that a certain bus wasn’t *their* bus.

    I know it will never happen, but it’s always been a pipe dream of mine to be able to see a real-time list of the order of buses in the tunnel. This would reduce some aggravation.

      1. It’s not a tall order either; the square thing at the beginning and end of each tube is a RFID reader. Each coach has an RFID tag on the front (left of the destination sign; near the wheelchair symbol) that has the bus’ number, route, operator ID, and more programmed in.

    1. Per your last comment, that would be great. I’m sure someone knows better than i do, but if there is someone “dispatching” the buses from CPS to Westlake SB and conversely from IDS NB, how hard would it be to have a simple interface where those folks update the departures. Its not as though there are changes in the order once the bus is in the tunnel! I think that would help on crowding, runners, and potentially platooning (operator: “oh, i shouldnt clear that 3rd 41 in a row, ill clear a 255 first).

    1. Could there be, in the tunnel? Or at least a rule about not having two of one route getting dispatched in the same platoon?

    1. Totally agree with this. Or, even better, give us cell coverage and/or wifi in the tunnel.

      Re: #4, why are two 41s running together hard to avoid? This is not my regular route, but I take it outbound every Wednesday evening to pick up my kids at Northgate by 7 and have been frustrated many a time due to delayed buses on this route. What makes it such a difficult route to keep on schedule?

      1. because it runs on I5 to/from Northgate … Northgate traffic sucks … and I5 sucks either way depending on the time of day.

        Add to that the fact that people will not yield right of way to buses … they get backed and bunched up.

      2. there is no point on route 41 between northgate transit center and downtown when traffic is required to yield to buses as there are no stops on the street and buses reenter traffic from ntc via a stop light.

  8. I’d like to submit this following comment for the next comment of the day post …

    The other day, I saw a D Line that said “Downtown Seattle.” Then later that day I saw a route 28 that just said “Downtown.” What’s up with that?

      1. I think the quoter was confused why one said “Downtown Seattle” and the other bus said only “Downtown”

    1. Maybe the RapidRide buses have more room on their headsigns, or are able to fit more letters in the same amount of space. The 28 frequently uses older buses with flippy dots.

    1. Based on the current revenues being 30-40% short, that will be in 2026, or 30 years after ST1 was downsized after a failed vote and passed in ’96.
      Ahhhh…. It seem like yesterday!

  9. “…there would still be delays caused by too many buses using Bay A. That list includes the 41, 71, 72, 73, 74, 76, 77, and 316. Given the rule that only two buses can be considered to be at a bay at once, and any bus further out must stop again at the bay, Bay A cannot handle the peak volume — 29 buses in one hour — unless they are spread perfectly evenly among the sixteen platoons that should be going through in that hour.”

    Then my solution would be to apply the 3rd-coach-must-stop-twice rule southbound (or inbound northbound) only, remove Bay B, and have the entire northbound platform serve as Bay A.

    1. I believe the two-bus-per-bay rule has to do with ADA. I’ve also pondered having the bay sign one bus-length back, to cleverly make the bay three bus lengths, but I don’t think that will fly.

  10. Why wait for runners in the tunnel? It’s not like there isn’t another bus coming and they are in a tunnel station, not standing in the rain. It would help if the drivers would tell riders to exit the rear doors. They don’t and it holds up loading. Metro needs to hold more driver training.

    1. “Why wait for runners in the tunnel?”

      Honestly, because doing so gives you a warm glow when people thank you profusely for waiting and often even send in commendations for doing so. I’ve heard many stories of REALLY SLOW drivers who wait for runners, help little old ladies with their grocery carts, and make sure everybody is seated before driving off. Metro’s culture to date has at least passively encouraged such behavior by emphasizing “service” over “schedule”. Metro’s culture is starting to change but it’s going to take a while to turn that battleship. Training for joint tunnel operations was the first time I heard training pushing us to move faster. From day one we’ve been told not to open our doors a second time. That said, many still do it. (See first sentence).

      Ideally, Metro will push us to be a little ruthless when we have to but also retain some of the old culture when warranted. (I’d argue there’s no need for me to leave a runner behind when I’m only a minute behind schedule with 15 minute headways)

  11. Another big issue I see is the boarding assistants can NOT effectively communicate with the drivers when they have finished loading rear door passengers. At least during heavy traffic times.

    I have witnessed the front of the bus being completely loaded and several seconds later the rear was as well. BUT it took the boarding assistant well over a minute to communicate with the driver to close the doors and move on….

Comments are closed.