I-405 BRT Corridor OptionsAmong the most interesting results of Sound Transit’s BRT studies was essentially no ridership benefit from “high investment” BRT options that spent more to dedicate right-of-way for buses.

I asked ST spokesman Geoff Patrick if the ridership model gave a bonus to routes that were more reliable. All models have to ignore certain effects, but omitting this one would have drastic implications for the computed value of bus capital investment.

The short answer is that the models do account for it, but perhaps not as much as someone like me might like. ST’s ridership model takes travel time as an input, and in general a bus with priority treatments will go faster. Moreover, the model assesses a higher ridership penalty for waiting time when compounded by uncertainty about the next arrival. However, specifics of the study corridors negate this time advantage.

In the case of 145th, the short distance means the potential time difference is small. Priority treatments like signal priority and queue jumps will make the usual time difference vs. a true bus lane negligible. And of course the vast majority of the route, on SR 522, will have dedicated lanes.

Meanwhile, on I-405 the investment is used to change the route, not make it faster. The high-investment scenario would add stops, and serve Tukwila Sounder and the planned dense neighborhood at Southcenter. These deviations increase end-to-end travel time enough to reduce ridership just as much as the new destinations increase it. Regrettably, the Southcenter plan doesn’t include Bus option “B2”, which would get buses out of Southcenter muck with the addition of a single on/off ramp.

This is somewhat ironic given Sound Transit’s primary purpose is to build 100% dedicated right of way, mostly light rail and commuter rail, without forcing every block of those alignments to justify those investments with a traffic study. On the other hand, new ROW is expensive, and the choice to use BRT is in practice a choice to cut costs and sacrifice reliability.

Although I don’t have a mathematical model to point to, one of the great things about truly dedicated ROW is that it reduces tail risk. When 145th st backs up because of an accident on I-5, one bus configuration will proceed with minimal delay and one will collapse in congestion. There’s no single answer for how much that peace of mind is worth. Some insist on grade separation; others will take the usually-good-enough priority treatments. For me, true dedicated right of way is the threshold between reliable transit, and not.

45 Replies to “Why Don’t ST Reliability Investments Score Higher?”

  1. I thought ST’s “primary purpose” was to build HCT, not necessarily to build “100% dedicated ROW”. While the two certainly overlap they aren’t necessarily the same thing.

    It sounds like the study properly accounted for speed and reliability improvements with BRT. It shouldn’t be that much of a surprise that the benefit is small. It often is small with BRT anyway.

    1. It’s all about the corridor. There is nothing special about this corridor — that’s the problem. Light rail would be no better, because you would have the same results. There are two different issues here. The first is that an improvement in reliability and speed to 145th is not that important. The speed difference is minimal. I find that questionable, but that is what they say. It would be the same if it was a light rail line running with or without signal priority.

      The second is a different story. You are basically changing the route to connect to different places. In the city, this leads to a network effect. For example, those from Northgate headed to downtown will certainly benefit from the detour to Roosevelt, UW and Capitol Hill. You will have plenty of people who get on at Northgate and don’t get off downtown. But in this case (as in most suburbs) you really don’t have that. The detours are simply a distraction. Those that benefit from the connections lose out because the primary destination (presumably downtown Bellevue) is slower. It is a common phenomenon, which is why applying one size fits all solutions are destined to fail.

    2. Dedicated ROW and grade separated ROW are two different things. The fastest ground transit option is Sounder, but it has lots of grade crossings so it isn’t grade separated.

  2. Dedicated lanes provide no additional benefit over shared lanes if you can actually make them reliable – it adds cost without additional functional benefit. What *does* matter is access, that is, does the bus have a reliable path that serves all its stations without weaving across traffic. Modeling is not as good at capturing things like having to weave across three lanes of general purpose traffic – it’s too detailed to be seen in the regional model they use, which looks only at average speeds (not variation) on a link depending on volume and capacity. Simulation is needed to understand more nuanced investment effects. But I’d start with a focus on access and a reliable pathway that avoids lane changes and allows station stops. That’s where there’s value to adding investment.

    1. The “making them reliable” part is the issue. You can’t make shared lanes reliable unless you can guarantee that accidents won’t happen (which is impossible) or you have a mechanism to pull stopped cars out of the road immediately (which also doesn’t exist, and there’s no place to put them — otherwise that place would already be a parking space).

      1. There was an accident on 145th east bound around 12th avenue the other morning, 2 cars rolled over according to tweets at the time. They closed 145th between 5th and 15th for about 3 hours luckily it was between rush hours. The only press was the incident and no follow up story.

      2. Trains have incidents too. The question is whether the right of way can be operated reliably, not what mode it is.

      3. Trains are a few large vehicles with professional operators. So there are fewer vehicles to break down or fill up the track, they get regular maintenance, and the drivers aren’t weaving back and forth and doing other disruptive things.

      4. Trains movements are also tightly controlled, via dispatching and strict adherence to operating rules.

  3. Regional models also look at parking availability and cost (which I think is treated similar to time). Parking is free and plentiful on the east side. Why wouldn’t you drive?

    On the other hand, models that treat time as always being equal are outdated. And I bet their model does. To me, if it takes 15 minutes longer on the bus but that means I get 45 minutes of smart phone time instead of 30 minutes of congested driving, I opt for the longer trip.

    1. My brother experimented with a 90 minute bus ride instead of a 60 minute drive and was amazed to discover he preferred the bus. Plus he said he’d had a live long desire to read Moby Dick and he was able to, all because he had all that bus time.

    2. That depends somehow on how productive one is able to be on a bus, as well as the busyness of one’s schedule. For instance, if the wait time at the bus stop already takes as long as driving to the destination, riding the bus suddenly looks a lot less attractive. Similar if you have to stand for the entire trip, or of the bus route is full of twists, turns, stops, and starts, making it impossible to read a book without getting motion sickness.

      This is one of the reasons why long-haul freeway express routes are so popular. The travel time (by all modes) is long enough so that the overhead of getting to the bus stop and waiting for the bus does not dominate the trip. And the motion of trip is smooth enough that one can actually read a book and not get motion sickness.

      1. But from what I’ve heard and read, we’re expected to not mind standing on the bus–aren’t the newer buses built to hold more people, but only because a greater percentage of people are standing?

  4. I would guess that people who can read on the bus comfortably have a dramatically different value of time than those who get carsick when someone else is driving… The models do account for a higher “cost” of time spent transferring and waiting than being on one’s way. They don’t generally add the higher cost of waiting when arrival times are uncertain or effective headways are longer due to bunching.

    1. Yes, I think that is the case. But a lot of people simply value their time more. For example, a parent trying to get home to pick up their kid from daycare. Yes, it is nice to read on the bus; but if it costs a bunch of extra money to do so, a lot of those parents will just drive.

    2. There’s really two cases here. People who have somewhere to be at a certain time, and people who don’t.

      If you have a tight schedule to keep, speed and reliability are king, and with regards to speed, it’s the total travel time that matters, while how much of it is spent walking, waiting, or riding doesn’t really matter all that much. If anything, it might be slightly better if a greater proportion of the trip is spent walking, since walking is 100% reliable and immune to traffic congestion; plus, the option exists to run portions of the walk to speed up the trip.

      For an example of this, on a recent airport trip, I was headed to the airport from the Columbia City, and was running late. Yet, I still opted in favor of Link, rather than Uber. Why? I figured that the wait times for Link and Uber at the Columbia City end would be roughly the same, as would in-vehicle travel times (Link through Tukwila is pretty fast), but by bypassing the long line of drop-off traffic at the airport, Link would be more reliable. Even the walk through the parking garage wasn’t much of a deterrent. I was traveling light and figured I could run from station to terminal in as little as 2 minutes.

      By contrast, if I don’t have a tight schedule to keep, I will often choose a more relaxing trip over the shortest trip. If the weather is nice, the more relaxing trip often ends being an hour long walk over a 20-30 minute bus ride.

  5. My bicycle is 99% reliable and I want transit to be also.

    Right now the E line is unpredictable.

    Not just in that it works but that I arrive at my destination as planned

  6. The reliability issue is representative of how staff only forecast AVERAGE conditions. It’s like forecasting climate as opposed to weather. Not until we forecast multiple scenarios can reliability be reasonably represented. Every other forecast that we have (snowfall, low temperatures, drug effectiveness, retirement funds projections, etc) widely reports variability. We should expect the same from forecasters.

    Not until we build models that simulate different land use forecast outcomes, different weather conditions ( both effects on supply and demand), special events, changes in working hours and a host of other things can we look at reliability in each of the possible conditions that vary from one day to the next.

    1. I’d also suggest sensitivity testing for parking charges, parking availability at both destinations and transit stations, employer transit subsidies, gas prices and HOT tolling rates.

  7. “When 145th st backs up because of an accident on I-5”

    Almost every day there’s an accident in Puget Sound somewhere. When I rode the 71/72/73X southbound in the PM rush (until I smarted up and started taking the 43 or 49 instead), it had unusually long slowdowns at Denny Way once a month, and 30+ minute slowdowns a few times a year. Grade-separated transit is well worth it to avoid these slowdowns that are more common than people remember.

    1. Right, but how much do you want to spend to make it more reliable? That is really the question. The great benefit of light rail to the U-District is added capacity and the connections that were previously horrible. It is not the connection from UW to downtown (a connection that is usually just fine) it is the connection from the UW to Capitol Hill. or Northgate to the UW, or Roosevelt to Capitol Hill, etc. There is no way we would have spent billions simply making the system more reliable. We will add capacity and critical connections, but reliability is just a side benefit.

      It really has to be considered on a case by case basis. Imagine we weren’t building a light rail line. Right now, the 41 is crowded, and sometimes unreliable. Would I like to make the route more reliable, or would I like to double the frequency on that, and every other bus in town. I would pick the latter every time.

      I’m not thrilled with Sound Transit’s modeling (that’s for sure). I think it is suspect, to say the least. But it is quite possible that there really aren’t that many people willing to take a bus along this route, no matter how fast. Spending huge amounts of money to make their trip just a bit better might be big waste of money.

    2. That’s a false choice though because when you build a dedicated ROW it simultaneously provides speed and reliability and room for frequency. If we started from the premise that there must be a network of reliable transit routes, then people would trust them more and be willing to use them. When Metro took a lot of hours during the various restorations and Prop 1, it was frustrating that buses didn’t get any frequenter. On the other hand, whenever I take the 131/132 they’re 5-15 minutes late, and that adds up over several days of round trips. The 14N used to do that: it would be ten minutes late so people would take the 43 because what if they didn’t and the 14 didn’t come? So reliability makes a substantial difference. At the same time, 145th may be too short for a transit lane to speed up the buses significantly; but conversely it wouldn’t cost a billion dollars to install transit lanes there.

      1. But every bus route could be improved. An example is the 131/132. Do we spend a bunch of money on that, or different bus routes. It is obvious that this set of bus routes really isn’t that good. That’s the problem. There are a ton of bus routes that are horribly unreliable, but way more important. If you made the 8 more reliable, it would make a huge difference to way more people than if you made this BRT route more reliable.

  8. How far out into the future do they look at their ridership vs time data? Some of the investments in I-405 for instance might not make a big difference in time today but when traffic volumes increase and other infrastructure improvements for high occupancy lanes are added I expect some of the “capital intensive” projects start to look pretty good.

    What I’d like to see in ST planning for buses on 405 is a complete system plan that works. Then you can work backward and see which piece of it it makes sense to invest in first. Of course this all has to be done in cooperation with WSDOT.

    1. Bernie,

      Exactly. The East Side of Lake Washington is going to continue to attract jobs and to a lesser extent, residents. There isn’t that much developable land left in the I-405 corridor, but that’s just going to mean that people commute from the funnel shaped territory north of Woodinville and south of Renton.

      So in 15 years when all lanes are congested for an additional two hours a day having separated lanes for buses will be worth a lot!

      1. This assumes the HOV lanes will stay 3+ forever. WSDOT is not supporting transit-only infrastructure.

      2. This assumes the HOV lanes will stay 3+ forever. WSDOT is not supporting transit-only infrastructure.

        First sentance, sometimes you just have to believe that not everything that could go wrong will. Otherwise you dig a hole in the sand and do nothing. Second sentance, clever wording to make it sound like WSDOT is anti-transit but the fact is transit only infrastructure has nothing to do with it. None of the freeway HOV lanes would make sense or even exist if it were for “transit only”. 3+ is pretty competitive with transit on a carbon emmission level. Of course the “gold standard” is living withing walking distance of your job but that’s not today’s reality. The real goal here is to stop building more GP lanes that only lead to more sprawl and will never “fix” or even have a medium term chance at eliminating conjestion.

        It’s been a while since stated but, we don’t have a capacity problem with our freeways. In fact we have an insane amount of excees capacity. What we have is a peak capcity issue but the large majority of hours in the day there’s a huge paved over part of our planet you could play basketball on.

      3. So you think that the current bill wiping out the HOT lanes will pass? If it does then the folks on the East Side will get what they asked for.

      4. So you think that the current bill wiping out the HOT lanes will pass

        If it did, which I don’t think it will, I’m confident our Governor would veto it. The State is on the hook for a huge chuck of Federal Funds which we have not way of “giving back” if we bail on the plan prematurely.

      5. WSDOT requires any “BRT” infrastructure in the I-405 ROW be able to accommodate auto drivers. Bus Only lanes, such as that currently operating on SR-522 can be narrower (10 ft.), since bus operators are specially trained.

        Amateurs (HOV drivers) require at least 11 foot lane widths, and breakdown room.

        And, IMO, – NO, I don’t trust the performance will be kept as high as could be achieved with exclusive an exclusive ROW.

        Of course, the ERC could provide that, but now it appears there will only be HSB there.

      6. WSDOT requires any “BRT” infrastructure in the I-405 ROW be able to accommodate auto drivers.

        What’s your point? There is nothing close to enough bus service for a dedicated lane. ST would never even consider the insane amount of money it would cost and even if was free it would be a terrible to create that amount of new pavement for a small amount of use over a tiny portion of the day. Transit/HOV lanes let buses basicly use for free WSDOT funded ROW.

        And, IMO, – NO, I don’t trust the performance will be kept as high as could be achieved with exclusive an exclusive ROW.

        False comparision. There is no exclusive ROW for transit on 405; not going to happen.

        Of course, the ERC could provide that, but now it appears there will only be HSB there.

        Another false comparison. The idea of transit on the ERC is so stupid I won’t even repeat why. Anyone that is opposing incremental improvements of transit via 405 by saying that money should all be spent on the ERC is either clueless or actively trying to poision a vote on a transit funding package.

      7. “The idea of transit on the ERC is so stupid …”

        That’s the spirit, ignore the analysis that’s been done.

        Express buses on I-405 have the same ridership as transit on the ERC. except that there is ZERO potential for TOD on the freeway.

        The ERC diverges from the freeway ROW in enough places that TOD development would work, if the municipalities don’t sabotage it by kowtowing to the ERC NIMBYs

        The secret would be to find the most bang for the buck, and there was an alternative that was cheaper, and had the same ridership numbers.

        Hey, I don’t want to rain on you Better Bus Parade, it will be the Eastside Sub-area voters that will decide.

      8. That’s the spirit, ignore the analysis that’s been done.

        The ERC diverges from the freeway ROW in enough places that TOD development would work,

        What analysis, where?

      9. Mike, there are lots of non-residential areas along the ERC. The buses or trains wouldn’t be stopping in those McMansion areas anyway.

      10. High CapacityTransit studies in the corridor began back in 1992.
        available at the Seattle Public Library:
        Regional Transit Project
        Eastside Commuter Rail Feasibility Study
        By Parsons Brinckerhoff/Kaiser Engineers Team

        Book – 1992
        R385.09797 P251R

        The I-405 Corridor Program – 1999 thru 2001
        Note, using the ERC never got to the Cost Benefit Analysis because the City of Renton, along with the Kennydale Neighborhood Association had the Executive Committee remove it from consideration.
        Light Rail (connecting the ‘density dots’) costing Big $$$ was compared to Express Freeway Bus costing less $.
        PSRC/Sound Transit ERC Commuter Rail study completed in 2009 concluded it was a viable project.
        WSDOT has a BRT White Paper on Freeway Buses.

        Look them up, read them.
        The 1992 study is only available in paper form.

        The only thing in the way of HCT on the ERC is those McMansions. It always has been.

  9. the choice to use BRT is in practice a choice to cut costs and sacrifice reliability.

    Say what? That is nonsense. It is exactly the opposite. Imagine if you said the following:

    The choice to use light rail is in practice a choice to cut costs and sacrifice reliability.

    That is just as ridiculous a statement, yet I’m sure I can find examples of cheap, unreliable light rail lines. Hmmm, there is one called the South Lake Union Streetcar — wow, I didn’t have to travel far to find just such an example.

    When you choose to build BRT you are spending more money to build something faster, more frequent and more reliable. Of course you are. Otherwise you can just run a regular express bus. Level boarding, off board payment, grade separation all cost money, yet all improve speed, frequency and reliability. That is the point.

  10. I’m assuming that these 405 BRT projections are based on WSDOT completing the extra 405 HOV lane between Renton and Bellevue. Once the extra HOV lane is complete, then bus becomes really competitive vs SOV drivers in the gp lane. It’d be nice if there was an easy connection to Tukwila station…but current Sounder riders can just continue transferring onto a bus to Bellevue from Kent Station (the 566/567).

    1. Kent,

      Commuters from Auburn or south would probably do better just to ride Sounder all the way to KSS and change to Link to Bellevue.

  11. Martin:

    ” Some insist on grade separation; others will take the usually-good-enough priority treatments. For me, true dedicated right of way is the threshold between reliable transit, and not.”

    Guess which local transit agency should have “Plerumque Satis!” as its motto. Because like all statements of solemn purpose, “Usually Good Enough” just doesn’t have the same “ring” as its Latin translation.

    “Mihi certum est dedicatum est inter limina dextra modo transit, et non morietur. ” Sadly, two words are max length for a Latin declaration, no matter how cute the girl across the Latin class aisle was in 1962.

    But by personal transit experience in every seat of a trapped bus including the left front one and every foot of the aisle, in the 18th century English of Old Virginia, Patrick Henry would have thundered: “Give me a transit lane secured by cannon, or leave me in a standing load between Everett and Seattle at 8AM!”

    Since the Fourth of July is only once a year and never celebrated at rush hour, Joint Base Lewis McChord has some canons they’re not using.


  12. At the time of the Revolution, spelling was a personal choice. But rush hour northbound I-5 at Northgate, the world’s worst choir is called for, because howitzers just won’t to the job.


  13. Look, the problem here isn’t about assessments of data, or even priorities. The problem what the Downtown Seattle Transit Tunnel has spent 25 shameful years proving:

    That the agency that decided to collect fares aboard Tunnel buses does not care if buses are slow. No matter how much damage that condition has inflicted on our flagship light rail service for going on six years. Which will look minor when U-LINK opens shortly.

    Compounded by the uncomplaining acquiescence of the agency whose colors are on the trains are held outside stations ’til those buses get moving. So it’s a fair choice whether to move DSTT bus service upstairs, or get both service modes straightened out at its present location.

    But if those Tunnel buses are still in the way of a subway full of trains anywhere near the ST3 vote, we’ll have a long time to work out Bus Rapid Transit. I apologize for the Revolutionary War crap, but can’t help but miss the Age of Reason.

    Mark Dublin

  14. I think the important thing about reliability is that it is something that public transit can get right. The roads are always going to suffer from traffic jams because increased capacity is going to bring increased use. By being reliable, even if not faster, a rail line can provide consistency that is valued by daily commuters and by people who generally just need to be somewhere at a predictable time. I wouldn’t expect reliability to be quickly rewarded, but over time I think that it is one of the primary benefits of public transit over personal vehicles.

    The basic question is how early do I have to leave in order to reach my destination at this specific time? If transit reliability gets me from home to work in one hour every morning then I can plan on that. And this might be preferable to using a vehicle which can get me to work in half an hour on a good day or an hour and half on a bad day.

  15. I kinda get the impression that a lot of the projects are intended to be toolboxes, rather than built as proposed. In this case, Sound Transit might be thinking of starting with the low-cost version and picking best pieces of the high-cost version to add to this.

Comments are closed.