Reece Martin (RMTransit) resurfaces with a discussion with Dr Jonathan English on how process issues affect the quality of transit projects. They call it “engineering issues” vs “phone call issues”. Engineering issues are straightforward technical problems, like building Link on a floating bridge or a tunnel in dubious soil. Phone call issues are those where a phone call could fix the problem or reduce costs: a call between politicans or agencies or with environmental-impact stakeholders. It might be more than one call or another communication method or a series of meetings meetings, but a phone call is a nice symbolic image.
They look at the goods and bads of Toronto’s transit: the recent east-west subway/light rail lines — Eglington Crosstown (#5), Finch West (#6), Sheppard East (#4) — platform screen doors or short fences, streetcars, too-close stop spacing, or insufficient street priority or signal priority. They look at why these happen, and how a “phone call” could have made these better. They also posit that social media played a key role in getting politicians to commit to fixing the notorious Finch West line slowness. And also that proposing to fix a problem throughout the city can overcome opposition more than proposing one of the same fixes at a time in a few individual locations.
Ideas they suggest include engaging with environmental-impact stakeholders in a neighborhood to come to a compromise or win-win deal, buying off impacted businesses in construction zones, or looking at how other countries can build the same thing much cheaper.
Compromise engagement could be a way to deal with the CID station alternatives issue, opponents of DSTT2, etc, rather than the board simply making a decision favoring one neighborhood faction and ignoring everything else including passengers’ needs.
Buying off impacted businesses could look like, if they demand lowering the line into an expensive tunnel for $1 billion, instead give 2% of that amount to the businesses to have a year-long vacation during construction. This may not be legally possible here due to a prohibition against giving public money to private businesses beyond EIS mitigation, or distributing it unequally to certain businesses. But it would be a way to build a better project at lower cost that opens sooner.
For comparisons with other countries they look at how Toronto couldn’t build full platform screen doors like Paris because it would cost too much, so instead they installed low metal fence panels, which you can still fall between or get pushed over onto the track. Part of the issue is that adding full-height screen doors can require a ventilation system upgrade, because older stations were built depending on whole-room air circulation. Still, Paris solved that problem and can retrofit stations at much lower cost, closer to the cost of Toronto’s lame fences. So maybe a “phone call” to Paris could allow Toronto’s subway authority to learn how Paris did it so that they could do it too.
The “throughout the city” issue is for something like stop diets. If you propose deleting one station or bus stop in one location, or three in different locations, then the people who would be negatively impacted overwhelm the public response and defeat the project. But if you propose to prune excessive stops citywide, then the larger number of people citywide who would benefit with faster transit trips would notice the issue and speak up to defend it, outnumbering those who complain. The net result is a better and more popular transit network citywide.
Another issue is TOD vs bus transfers. They make a case that a couple large TOD towers have only a thousand residents, only some of whom would use transit. A popular bus route brings tens of thousands of people to the stop, and all of them are already on transit, so more of them are likely to transit to the subway to all the destinations the subway goes to. In the case of Toronto’s streetcars at their peak (before their service degraded with slowdowns), it was hundreds of thousands of people per line. So agencies should prioritize connecting to popular bus routes and having a good bus-feeder network. Applicability to Link includes DSTT2’s bad downtown transfers, the lost midtown transfer to RapidRide G at Madison Street, dropping Avalon station (route 21), the bad Overlake Village station location (and Metro not rerouting the 226 to Spring Blvd near two Link stations), Capitol Hill station being at John Street instead of Pine Street (the primary east-west bus corridor), and probably other station issues I’ve forgotten.
They also fault Sheppard East for being too short (five stations). They say it could be extended a few stations to a busy bus route, and that would significantly increase ridership. Conversely they praise Eglington Crosstown as being decent sized and fast so you can travel across a large percent of the city in twenty minutes.
Note that this is “citywide”: in an area as dense or denser than the inner half of Seattle. Toronto is dense over a much larger area than Seattle, and all these subway lines and streetcars are within that dense area for the most part. When you get out beyond say Northgate or Lynnwood, there are fewer people, bus ridership is lower, and the problem of “a few small dense islands in a sea of low density” becomes more acute. Those all make the issue less applicable. The closest comparison to Sheppard East may be extending Lynnwood Link to Alderwood Mall or Ash Way (as earlier proposals for ST2 would have done, although that’s not really about busy bus routes. A better comparison would be if Link ended at U-District and there were a proposal to extend it to Roosevelt or Northgate: that would bring in the 45 and 62 and all the Northgate routes. That would be a significant benefit, although none of them are probably as busy as the Toronto crosstown route, and none are BRT (RapidRide) lines. But still, extending a U-District terminus to Northgate would be a significant benefit. That doesn’t work for extending Federal Way to Fife.
This is a semi-open thread, for process and network issues, or transit during the World Cup month. A full open thread will be tomorrow, or the last article before this.
In the meantime, can we get a RapidRide stop in Fremont? or more bus routes along Stone Way? I wonder what the difficulty is in adding a RapidRide E stop in Fremont.
A route 40 upgrade is currently under construction to give it almost-RapidRide-like service. (June 16 update.) It was one of seven routes chosen for RapidRide in the Move Seattle levy in the 2015, but overoptimistic budgeting prevented it from happening. The current upgrade is a compromise to at least get something better.
(The full list of Move Seattle RapidRide lines was the G/120, H/120, J/70, R/7, 40, 44, 48, and 62. (That’s eight. Maybe the G was outside Move Seattle, or the H was added afterward.) The first two are finished, the third is under construction, the fourth is awaiting full funding, and the last four were canceled.)
Several of us have been contemplating a RapidRide E station approximately above the Troll that would have an elevator down to Fremont. That would require Seattle or Metro to recognize it as a priority and fund it. They have so many other high-priority transit needs they have limited money for that I doubt they’ll recognize it. But if the 40 is upgraded, and if it gets more incremental upgrades later, then maybe we won’t need an E Fremont station as much.
Interesting, why was the 62 cancelled? It seems like Stone Way could benefit from more transit
Basically, the RapidRide projects grew from being “let’s build fancy bus stations at major stops with real time arrivals screens and curbside ORCA readers, paint some strategic bus lanes, and run the bus every 10 minutes” (see RapidRide A through E), into much more complicated “let’s rebuild the entire roadway and subsurface utility infrastructure with transit grant dollars and also do the other stuff”-type projects (see RapidRide G, H, I, and J lines). It turns out the latter version costs way more than the former, and the capital funds ran out after COVID. So, Metro reprioritized the unbuilt RapidRide lines and cancelled a bunch that were originally planned but would end up being too expensive (https://seattletransitblog.com/2024/07/23/rapidride-future-and-prioritization/). This was in the midst of folks realizing that RapidRide had become more of a road reconstruction program than a BRT program and advocating for transit dollars to go toward transit improvements, not vaguely associated roadway and utility upgrade work (https://seattletransitblog.com/2024/03/03/no-more-rapidride/).
Ideally, we’d go back to doing the basic version of RapidRide and start putting up nice stations and red paint instead of committing to full street rebuilds. But that probably requires some backtracking on how the projects are perceived among the various agencies.
Because Kubly overpromised what the Move Seattle budget could afford, so it had to scale back. I think there was also some recognition that route 62 was maybe not the best corridor for RapidRide so the whole concept had been a bit fanciful. The 62 is definitely a strong route with a lot of overlapping trips between downtown, SLU, Fremont, Stone Way, Greenlake, Roosevelt, 65th and Sand Point. But it’s still an L-shaped route, so it’s maybe not as high-ridership as one that goes north-south to the edge of the city, or east-west to Ballard.
RapidRide isn’t the only way to improve service on Stone Way. What does Stone Way need? More frequency? The 62 is already 15-minute frequent full time.
There’s an on-and-off-plan to straighten out the 62 so that it would continue east on 45th Latona, north to 65th, then east. That would speed up trips between Stone Way and Roosevelt, and thus the entire corridor. It was in the RapidRide plan and in Metro Connects, and it would have been in the Northgate Link restructure, except that Latona Avenue N needs to be hardened to handle the weight of so many RapidRide buses, and SDOT has not yet scheduled that or budgeted it, so the restructure is waiting for that.
“and run the bus every 10 minutes”
The RapidRide minimum is every 15 minutes. The C, D, E, G, and H are above the minimum because they’re in such high-ridership areas. They haven’t always been. The C and D launched with 15-minute service. They went up to 10 minutes daytime in the mid 2010s, then back to 15 minutes in the covid recession, and now they’re up to 10 minutes again.
The G, H, and J are really SDOT projects. Metro didn’t want the G in its current corridor; it wanted a Broadway-Madison route. It thought that would better match high-volume travel demand. But SDOT insisted on an all-Madison route to 23rd, MLK, or Madison Park, and strong-armed Metro into acquiescing to it. Since SDOT was paying for it, SDOT got to choose the alignment.
The H was accelerated by Seattle for equity reasons, to get frequent/fast transit into lower-income eastern West Seattle in the high-ridership, high-transit-dependent Delridge corridor.
The J was a Metro corridor but then Metro was going to drop it due to lack of money, and Seattle stepped in to pick it up and make sure it gets completed.
Right, which is the main problem with the modern evolution of RapidRide into predominantly roadwork projects instead of predominantly transit service improvements. I think SDOT realized RapidRide could be used to win FTA matching grants to help pay for major roadway and utility upgrades, so that’s what you see with G (complete rebuild of most of Madison), H (rebuild of most of Delridge with more street trees), and J (complete rebuild of Eastlake with protected bike lanes). The first few RapidRide lines stuck with strategic improvements, like bus bulbs, traffic signal priority, and painted lanes (https://seattletransitblog.com/2009/01/21/first-details-on-ballard-rapidride/)
You make it sound like Madison Street wasn’t an important transit corridor in its own right, or that the center lanes aren’t a major transit improvement, that it was just to get transit-grant money for utility projects. How were Madison, Delridge, and Eastlake not the best transit corridors for the next RapidRide lines? First Hill has the hospitals and highrises and Madison Street got a lot of growth in the 2010s, and First Hill desperately needed a better transit connection to the rest of the city and region. Delridge is a high-ridership, lower-income corridor that West Seattle Link should probably have served but didn’t. Eastlake is the primary local corridor between the U-District and downtown and is high-ridership and has SLU. All these seem like justifiable choices for the next RapidRide corridors.
And if the whole street or utilities are upgraded too, how is that a problem?If they needed to be done anyway, you might as well do them all at once to minimize disruption and cost. Isn’t SPU paying for the cost of the water main upgrade?
“J (complete rebuild of Eastlake with protected bike lanes)”
That I see as the city having inconsistent priorities that change depending on who it last talked to. So it passes a levy to get a first-rate transit corridor, but then bicycle interests convince the city to put a first-rate cycletrack, and that demotes the transit because there’s not enough room for first-rate of both. We’ve seen that now with both Eastlake and Broadway.
I’m not saying they weren’t, and I have no idea where you’re getting that from. I’m saying that the early RapidRide projects were cheap and fast to implement because they had very limited scope.
Because it makes the project slower and more expensive! The later RapidRide projects incorporated much more significant heavy civil construction elements, and got federal grant dollars to help pay for it, but it still made cost more out of local pockets. I don’t particularly care whether it was appropriate to include that work in those projects or not, but it’s a plain fact that including those roadway and utility rebuild elements jacked up the cost and stretched out the construction timelines to the point Metro began assuming future projects would be just as expensive. As a result, many of the planned RapidRide lines are now indefinitely deferred. If the goal is to spend transit dollars improving transit service, then RapidRide has grown into a very inefficient way to do that. That’s the thesis of Fesler’s “No More RapidRide” Op-Ed, and I generally agree with it.
You mean we could implement all the identified RapidRide candidates near-term with existing funds if it were determined that all these other things wouldn’t be needed in them? Metro or Seattle has never said that, and Metro has never said it’s deferring them because so many other things are being added that it becomes too expensive. And Fesler’s article didn’t say that either. They all gave the impression that RapidRide conversion was just intrinsically too expensive to pursue them faster. E.g., to buy the red buses, install the fiber-optic cable for the ORCA readers and next-arrival displays, build the transit lanes and bus bulbs, etc.
If it is true that a bunch of non-transit features are holding back RapidRide projects, then we should raise this with the politicians, and ask them to identify which corridors don’t need those, and build those lines next. And ask them to identify how much of the cost is directly for RapidRide infrastructure, vs how much is for other things, so that we could see it transparently and make an informed decision on those corridors.
But we can’t get out of some street costs. Adding transit-priority lanes requires rechannelizing the other lanes, and the street may have bumps and potholes that need to be fixed for RapidRide to function well.
Mike, Fesler was explicit in his criticism of cost bloat in RapidRide projects.
Quote:
It’s the roadwork and full-width redesigns that are holding RapidRide back. We basically need to decide if we want 1-2 RapidRide G, H, or I lines, or 5-6 RapidRide A-F lines.
… that’s what folks have been doing! Fesler started it in 2024, and Metro’s been dealing with other budget issues in the meantime. I fully expect there to be pressure to reduce costs of the RapidRide projects proposed under the new “Next Stop” plan being presented to King County’s Regional Transit Committee today.
My contention in my No More RappidRide article, which I wrote in response to Fesler’s, was the same as what I thought Fesler was saying. Namely, we don’t need full-on RapidRide conversion as much as we need incremental improvements to increase the speed, frequency, and reliability of routes.
The biggest thing RapidRide gives us is guaranteed 15-minute frequency full time until 10pm every day. That has made me more willing to go to West Seattle, Aurora, or Ballard in the evening or weekend than I did before, because I don’t have to work around half-hourly buses. That benefits residents in those areas and brings customers to those businesses. I used to live in Ballard, and one of the reasons I left was the travel-time overhead to get anywhere else, and the fact that most of my non-work activities where somewhere else. The infrequency was a secondary issue, but since it had always been like that I thought nothing could be done about it.
You can have guaranteed frequency without RapidRide, but that requires a guarantee. Frequency went up on the 40 and 5 and other routes, but Metro made no long-term guarantee that they would always remain frequent or that it wouldn’t cut them in a recession. That’s the difference between RapidRide availability and other routes. Seattle’s Frequent Transit network says it aims to make most routes 15-minute frequent, but it doesn’t actually say all evening, and we don’t know now when the FTN will be fully implemented, or that it will always remain so after that.
The second biggest thing RapidRide provides is next-arrival displays. Average passengers hate waiting long periods more than the hate practically anything else, and the displays mitigate that. It’s physically better to know how many minutes you have to wait, you can plan interim activities accordingly (like getting out a book, making a phone call, popping into a convenience store) and know how much time you have to get ready to step on the bus (an issue during the pandemic when I had to put my mask on, glasses on, gather my shopping bags, and put my book in my backpack all before stepping into the bus, all without letting my glasses fall off), and it’s a psychological relief to see the minutes and visualize them going down one by one.
The third biggest thing RapidRide provides is more standing room inside and more space to maneuver getting on and off.
The other things — red buses, off-board payment — aren’t as important. Usually there’s only one or two people at each stop, so it doesn’t matter as much whether they pay offboard or onboard compared to if it’s a large crowd. But large crowds occur only downtown and at major transfer points like 46th, and even those don’t happen much anymore after the pandemic.
So just build the transit-priority lanes, make buses more frequent, and guarantee frequency long-term. Don’t spend money on red buses and off-board payment — and don’t use transit money for relocating utilities or or changing the rest of the street — if it gets in the way of improving more bus routes in more areas.
However, if the powers that be want to do RapidRide conversions, it’s better to do them than not do them. RapidRide I, J, K, and R will genuinely improve transit mobility in those areas, making a better quality of life for those who live there or travel there. We’ve seen it with RapidRide A-H. It may not be the best that’s possible, but it’s substantially better, and it’s better to have that than nothing.
I mean, we should be converting all routes to “Rapid” Ride in some capacity.
Real rapid ride would be like a BRT, with full bus lanes, actual signal priority so it blow through traffic without stopping anywhere except the bus stops.
Yeah, that is really a different beast. It basically means busways for a significant part of your city. The closest we ever came to that was the bus tunnel and the freeway connections. But those were spotty (e. g. it didn’t work from the north if the express lanes were going the wrong way).
But even then the focus should be on the busway, not that the buses on it are necessarily “BRT”.
Marco Chitti wrote an excellent critique of the whole BRT idea. It is well worth reading in its entirety (don’t base your opinion on my summary). As I see it, there is really only one reason for a city to create a “BRT” route: government funding. Otherwise you can do the same thing without all the ridiculous amount of paperwork. Just as BRT is not an all-or-nothing alteration, neither should we focus only on individual corridors. We should make similar fixes in multiple areas. We should prioritize fixes based on how much they improve the overall network not on whether a route is now “special”. For example, improving speeds over the Fremont Bridge could improve four current routes and more in the future. These buses carry more riders than any individual RapidRide route. Other than funding, the only major distinction is branding. The value of this is minimal and ultimately sends the wrong message (“all of our buses suck — except these special ones”).
The same thing is true with RapidRide. We can have a stop diet, BAT lanes, bus lanes, frequent service, off-board payment, rear-boarding and good bus stops with or without the RapidRide designation. If anything, many of these improvements make more sense when you don’t have a single RapidRide route. It is great that you can wait at certain “stations” on Aurora and read a kiosk telling you when the RapidRide E will be there. But it doesn’t change anything. In contrast, a similar kiosk outside a Northgate Station bus stop (serving several regular buses) makes all the difference in the world. For example, I prefer the 348 but the 61 will do. If the 61 shows up as I reach the street it is very handy to look at the kiosk and see if the 348 will be there soon or not.
Mike mentioned the commitment to good frequency that comes from RapidRide. If you ride that bus it is clearly a good thing. But if you step out and consider the overall network it isn’t. That service has to come from somewhere. It means that some other bus is getting less service simply because the RapidRide route has a special designation. Likewise, even making relatively small changes to RapidRide requires extra work as it may mean several more buses that are a different color. You want your transit system to be as flexible as possible. Obviously there are some improvements that are permanent in nature but the relative permanence of both the routing and frequency of RapidRide routes is a clear negative.
Thus it really comes down to funding. The federal government is more likely to chip in for a corridor-based project instead of a hodgepodge of small improvements spread out over the city even if the latter is a much better value. I can sympathize with the agency but ultimately it is a flaw with the funding and the wrong way to approach transit improvements (as Marco Chitti wrote).
“improving speeds over the Fremont Bridge could improve four current routes and more in the future. These buses carry more riders than any individual RapidRide route.”
RapidRide A-H are now 25% of Metro’s total ridership. So if one of those routes were on the Fremont Bridge, it might very well carry as many people as three of the other routes.
It turns out that when you have guaranteed frequency, and next-arrival displays, and station signs that say “E Line to Aurora Village”, and a fat line on the map, and marketing for those routes, people flock to them, even if they have to walk a little further, or further than a regular route they could also take, and ridership keeps increasing every year.
“We should prioritize fixes based on how much they improve the overall network not on whether a route is now “special”.”
The routes were chosen because they’re the most critical corridors, and one in each part of the city, so that people can at least get to that part, and then their remaining problem is smaller.
“Mike mentioned the commitment to good frequency that comes from RapidRide… But if you step out and consider the overall network it isn’t. That service has to come from somewhere.”
Only if you consider transit a zero-sum game. The problem is the governments are doing it backwards. The right way is to determine what we fully need, then calculate how much it costs, then make a plan to get from here to there, and have steps and milestones so that the public can see the city getting closer to it every year, and see how much of it has been accomplished.
Instead they start with the status quo service hours, and then give excuses why they won’t get to really good service, and won’t even set an adequate goal, and then they add a few service hours and won’t do any more, and we remain stuck in this transit network that only meets half the people’s needs half the time.
The right way to implement RapidRide is to add service hours for it, so that you’re not taking hours from other routes.
But when you have a zero-sum game, the RapidRide corridors are already a decision where the biggest needs are, so adding hours to RapidRide lines or adding hours in the most-needed corridors, are the same thing.
People have been suggesting a bus stop for Fremont from before there was a RapidRide E. It is just one of those projects that have never advanced. There are a lot of similar projects, some of which were part of Kubly’s “Move Seattle” project (which as Mike pointed out, ran out of money). For example, Jackson street could use additional right-of-way, as mentioned here. That fix implied it would be RapidRide which may have contributed to the problem. Then again, the 62 should be modified and that has nothing to do with RapidRide. These are just SDOT projects that pile up while we raise money to eventually build light rail from West Seattle to SoDo.
The reason things look so bad with frequency is we’ve been in a de facto transit recession since 2020. First there was policies with a 25% capacity cap and “Essential Trips Only”, then Seattle’s STM rate went down, then the driver shortage caused Metro not to run all its intended service, and the driver shortage also swallowed intended expansions like 15-minute evening/weekend service on the 550, 15-minute service on the 522 when it moved to Roosevelt, and route 61 initially.
Next year when Metro fully restores the driver-shortage hours, and if the STM increase hopefully passes, we’ll get closer to a 2019 situation, and then people will have more mobility choices and feel better and more optimistic. I don’t know if SDOT will return the 65/67 to 10-minute daytime frequency, but if it does then we’ll have another really good route pair in northeast Seattle, like we did in 2016-2019, and everyone who uses that route will be glad of it.
This is mostly in Seattle. The county as a whole and the other cities don’t have anything comparable to the STM boost, so their 2019 service level wasn’t that enhanced and their 2027 service level won’t be either. But it’s kind of their own fault for not doing anything about it either then or now.