Digging Into TriMet’s Proposed Service Changes: NW Edition

It’s time once more to dive into TriMet’s proposed service changes. This time I will look at the areas where the vast majority of actual bus service cuts will be concentrated: NW Portland and St. Johns. First, here are the maps of current and proposed service:



First of all, the big service cut. TriMet is proposing to completely cut the NW portion of the 17.

Currently the 17 runs from the downtown transit mall through the NW on Everett/Glisan and 21st to Montgomery Park, with every other trip continuing to St. Johns, Linnton, and Sauvie Island. Service between Montgomery Park and downtown is pretty substantial on weekdays, with 53 trips per day in each direction and roughly 20-minute frequency all day. Weekend service is very weak, however, with only 20 trips on Saturdays and no service at all on Sundays.

The current proposal is to have the 17 from Holgate stay on the transit mall and turn around at Union Station, cutting the NW portion completely. Service from Montgomery Park via 21st and Everett/Glisan would be replaced by the 77, but that is a crosstown bus that heads east across the Steel Bridge rather than downtown, so current 17 riders headed downtown would be left with three main options: Take the 15 on NW 23rd, take the Streetcar on Lovejoy/Northrup, or take the 77 and transfer to buses or MAX at the transit mall.

None of these are completely terrible options, considering the frequencies involved. The Streetcar is very frequent (74 trips per day) and the 15 and 77 are close behind (60 and 53 trips per day, respectively). Transfers from the 77 to service on the transit mall should be pretty quick most of the time. All these would require additional walking and waiting time, however, and the frequency of remaining services does not make for the fact that the NW is losing a large amount of service with the loss of the 17.

TriMet makes the case that service was never adjusted in response to the Streetcar, which is true enough. They point out that the 17 and the Streetcar both link PSU to the NW, so they considered redundant service. This is a tricky case, because although they share those two points, they take two quite different paths between them, arguably serving different markets. For example, the 17 travels along 5th and 6th through the heart of downtown, while the Streetcar travels along 10th and 11th, a corridor still relatively devoid of major destinations or employment. Another problem is that the Streetcar is simply slower. A trip from PSU to NW 21st/Northrup on the 17 takes about 18 minutes. The same trip on Streetcar takes 25 minutes. Given all this, I can see why folks in the NW might not see the Streetcar as an adequate replacement.

While people living in the inner NW (or “Alphabet District” as it is called by absolutely no one) at least have other options, the greater impact of this cut will be felt by transit riders in St Johns, Linnton, and Sauvie Island. They will lose their all-day service to downtown through the NW, leaving them with peak-only weekday service via the 16, which runs along NW Front Ave. The section of the 16 out past St Johns in the N quadrant will be replaced by a vaguely-defined “shuttle service.”

I am somewhat sympathetic to TriMet here, because this among the most expensive types of service to run, with buses running between small, low-density neighborhoods with very little between them. NW St Helens may be the most direct route from St Johns to downtown, but with so much essentially empty space between them there is little chance for the kind of ridership that makes for efficient bus service. That said, it would be a shame to eliminate the all-day service that many people in these areas have come to rely on.

I’m going to try something new here, to bring some data into the discussion. I present below a ridership chart for route 17 (click to embiggen):


This is a chart of stop-level ridership data from Route 17 southbound on weekdays from Sauvie Island to Holgate &134th. Stops are on the horizontal axis, and average passenger load is on the vertical axis. The length of the white bars represent the average number of people getting on at each stop, the black bars show the average number getting off at each stop, and the point where they meet is the average passenger load at each stop. Basically, by reading the chart left to right, you can see roughly how full the buses are by looking at the height of the bars, and you can see the ons and offs by looking at the length of the bars. I hope that makes sense. (Full credit goes to Bruce Nourish at Seattle Transit Blog, who invented this type of chart)

So what does this chart tell us? Well, it’s a good one because it contains examples of what a well-performing route looks like and what a poorly-performing route looks like.

As you can see, from Sauvie Island to St Johns there is almost nobody riding, with about 3 people per bus on average and virtually nobody using the intermediate stops. Enough people get on at St Johns to boost average load to about 8, and virtually no one gets on or off until closer to Montgomery Park. These flat lines are a clear example of an expensive route that exists to take a small number of people from point A to point B.

The rest of the route, especially once we get to the Holgate segment, is a good example of a well-functioning route. Most stops see plenty of ons and offs. This kind of rider churn shows that the route is serving lots of origin-destination pairs rather than just one or two. Average load gets up to about 20 and only slowly drops down, which means there is substantial demand all the way to the end of the line. This pattern shows a productive route that exists to take a large number of people from point A to B and B to C and A to C and C to E and B to D, etc. etc.

What this chart tells me is that if we have to cut service, it makes sense to cut the Sauvie Island portion that is barely used. That’s not to say there shouldn’t be any service, but perhaps it would be better to run far fewer trips than currently. Service from Montgomery Park to downtown, however, is quite strong and should be served in some way.

Well, that covers the service cuts. Is there any good news to be found? Quite a bit, I would argue. Much like the changes proposed for the NE, TriMet is using the unfortunate need to cut service as an opportunity to also make long-overdue improvements to the structure of the system.

By moving the 77 from Lovejoy/Northrup to Everett/Glisan, TriMet will vastly simplify the transit system in the NW.

Currently we have a confusing tangle of overlapping routes in the area, with the 77 as the worst offender. After it gets off the Steel Bridge, rather than continuing west on Glisan it makes a series of turns past Union Station, parallels the Streetcar along Lovejoy/Northrup, and for no discernible reason runs northbound on 23rd and southbound on 25th (even though the 15 runs two-way on 23rd). The 17, meanwhile, has to make a difficult right-left-left turning maneuver to get from Broadway northbound to Glisan westbound. Montgomery Park, hardly a high-density transit market, finds itself served by three relatively frequent bus lines simply because it’s a convenient place to turn a bus around.

By removing the 17 and moving the 77, the structure is very straightforward, with fewer turns and overlaps. The 77 will get off the Steel Bridge and simply continue west on Glisan all the way to 21st. Lovejoy/Northrup will be the sole domain of Streetcar, 23rd will be sole domain of the 15, and the unnecessary service on 25th will go away. The only remaining overlapping service will be at Montgomery Park, but at least these will be two buses, the 15 and 77, that go to completely different places.

Overall this new structure will give the NW something closer to the clean grid typical of SE, NE, and N Portland. Such a grid does require more connections, but it also normally allows more frequency because buses are being used more efficiently. Ah, and there’s the rub. Because this is being done in the context of a service cut, many of the benefits of a grid may not be apparent. In an ideal restructuring, the 17 would still be cut, but the service hours would be reinvested in the remaining bus lines. If the 15, 77, and Streetcar could be dramatically boosted in frequency, most people would experience shorter travel times with less waiting, even people forced to transfer between the 77 and MAX or bus lines at the transit mall.

Here’s my pitch to TriMet and to anyone reading this. Let’s agree to go through with the smart restructuring components of the budget proposal regardless, but if any extra funds are found over the next year, reinvest it primarily in these NW bus lines where most of the proposed cuts are occurring. Only $2 million of the $17 million projected shortfall is coming from bus service cuts, and given the high amount of uncertainty in projected payroll tax revenue and federal transit grants, it is not impossible to imagine $2 million or more could be found.

If the shortfall ends not being as bad as expected, I advocate continuing with most of the proposed cuts and restructuring, and investing the extra money in several ways:

  • Increase frequency on remaining NW transit lines (15, 77, Streetcar)
  • Extend the 77 to St Johns and Sauvie Island to restore all-day service
  • Restore the Frequent Service Bus Network
  • Restore MAX off-peak frequency to every 15 minutes

I would also add that TriMet and the city should invest in the area from Everett to Glisan and 5th to Broadway to make that area a more safe and attractive place to transfer. Currently I wouldn’t blame people for being wary of having to transfer to and from the 77 given the state of that area, especially the constant open drug-dealing around the Greyhound station. I am all for relying more on connections in a grid-based transit system, but if we are to do so we must make sure those connections are easy and safe.

28 Responses to Digging Into TriMet’s Proposed Service Changes: NW Edition