Providence: Busy Versus Frequent Buses
While trying to come up with a good proposal for upgraded buses or streetcars in Providence, I tried to base route decisions on RIPTA’s most frequent buses. But as it turns out, there’s a substantial difference between the most frequent and the busiest routes, and existing policies toward investment do not reward high ridership at all.
By far the two busiest lines in the state are routes 11 (Broad Street), with 6,500 weekday riders, and 99 (North Main to Pawtucket), with 5,200. Those are also the two most frequent, with 10-minute peak and midday service, and are usually interlined. This is the only case in which frequency matches traffic: of the next batch of busiest routes – 20, 22, 56, and 60, each with about 3,000 weekday riders – only the 56 has 15-minute off-peak service, the rest ranging from 22 to 35, with the 20 and 22 having 22-23-minute frequency even at the peak. Several less busy lines have 20-minute all-day service, and the frequent network, which uses a 20-minute weekday off-peak standard, looks different from the highest-traffic network.
However, previous and proposed development-oriented transit, including the fake trolleys and now the streetcar, avoid even the 11 and 99. The fake trolleys are distinguished in branding, 20-minute frequency even on weekends and in the evenings, and consistent interlining across Kennedy Plaza. The 92 fake trolley runs from the East Side to Federal Hill without changing its number, but regular buses, including the 11 and 99, change their route number at Kennedy Plaza, and that’s if there’s a consistent route they interline with at all. (When Jef Nickerson pressed RIPTA on this issue, RIPTA said it wants to preserve flexibility.) Likewise, the streetcar is a city-center circulator, and ideas for where to extend it afterward avoid Broad Street and North Main; local transit activists I have talked to believe the preference is for Broadway, a wide street hosting two routes (27, 28) that have 4,500 weekday riders between them, still less than Broad. (The alternative route in the same direction, Westminster, has 3,500 on its two buses, but the difference comes from the routes’ respective tails west of Olneyville Square, and the segments along Westminster and Broadway look about even.)
This is not to say that the state spurns the busiest routes. After the previous Governor vetoed it six times, Governor Chafee recently signed a bill to provide bus signal priority on the busiest lines. The brand for this is called rapid bus. At best, this shows the state thinks that rich people on the East Side and the Federal Hill gentrifiers, and soon the Jewelry District gentrifiers, prefer to ride a service that’s not called a bus, even if it is one. At worst, it points to skewed priorities: the streetcar is explicitly a development tool, and much more expensive than clearly posting schedules at the top end of the bus tunnel and rearranging schedules to provide constant headways within it.
A related issue is the ability to railstitute bus routes. Among all the busy routes, route 11 is among the hardest to replace with commuter rail. Peter Brassard’s urban shuttle proposal and my Woonsocket regional rail proposal use existing railroad lines. Arguably, this could take over the longer-distance functions of the 99, whose demand primarily comes from Pawtucket rather than North Main in Providence. However, the 11 is not paralleled by any rail line. This makes it the most important corridor for any upgrade. Alternative routes, such as continuing the existing streetcar proposal farther south, do not capture the local demand on Broad, which is of moderate intensity everywhere along the corridor. The distribution of demand on Broad is linear, which is less the case for other routes, which connect various anchors spaced farther apart.
It’s not normal for the relationship between traffic and frequency to be so weak. (In New York, busy routes that aren’t frequent by a 10-minute standard are the exception, and are very close to making the cut, e.g. the B8 and Bx39 run sometimes every 10 minutes midday and sometimes every 12). RIPTA needs to be asking itself why some routes are overserved and others are underserved.
But more importantly, the city and the state need to ask themselves why they’re building special branding as not-a-regular-bus around routes that aren’t even the ones that most need it. The fake trolleys get emphasized and specially colored on the map. It’s RIPTA’s fault that the interlined buses aren’t consistently signed, but all of the investment decisions are on the city and the state. Even if it’s necessary to build a streetcar to the Jewelry District and the hospital, why not say that pending additional funds the city will extend it toward and then along Broad? The alignment wouldn’t be any more awkward than that already proposed, and it would only miss a relatively short segment of Broad.
Pedestrian Observations 
Incidentally, it’s just Providence’s bad luck that the top bus route is far from any alignment that could be used for faster trains than streetcars. In New Haven I can’t find ridership data, but the most frequent route, D-Dixwell, parallels a defunct canal that’s now a hiking trail. (The idea of making walkable streets and having people hike on them instead is too radical, or something.)
Alon, this is why I like Peter’s proposed streetcar system (https://maps.google.com/maps/ms?msid=217434767782332614082.0004beea228495372a66c&msa=0). It has some rather ambitious extensions and perhaps some of them, particularly to the north, could be eliminated with a simultaneous plan to build the major urban infill shuttles. I think this could complement some of the urban shuttle stops that lie on a populated, but not heavily job-oriented corridor around the perimeter of Providence. It also builds on the currently proposed core connector in some ways that make the proposed route almost seem well planned.
I could easily see building the currently proposed core connector, followed by a southern extension down Broad Street to 95 and a westward extension (the red line) to the Manton Avenue/Plainfield split. From there, ridership and demand could guide additional extension along the “red” line to Manton. The Smith Street “blue line” extension and the Hope Street “red” line extensions could follow (although not likely initially to the proposed terminus). Initial extension on the “green” line north can wait until the urban infill station you talk about at the Providence Plaza Station becomes a reality, with further points north being serviced based on ridership.
Essentially, this map provides the “railstitution” of popular buslines, incorporates the currently proposed core connector in a coherent way, and even could connect sensibly to additional urban shuttle along the NEC rail lines, forming the “wheel” to these spokes. Integrated fairs and timings could actually lead to easy travel from outside of Providence into various parts of the urban core.
The light blue Line 2 Alt looks strange. What’s the reason for these detours? Serving Providence College, which is only 200m from Smith Street? IMHO, if there is any dent in the straight line, Chalkstone Ave. with the 2 Medical Centers would deserve it more.
Any funding commitments for service on the frequent routes?
The sheer ridership density of the 11 makes it quite surprising that it’s not the first place they’d look to improve service by running a streetcar (or second place, after the tunnel). A 3.6 mile line with 6500 riders a day is not too bad, and assuming streetcar running times are like the bus, they’d only need six cars to operate the line, and there are even a couple suitable locations for depots. Also, the existing 10 minute frequency is a solid base to build on, because something like a streetcar really does need that level of frequency to really be useful and make good use of the fixed investment. Plus, it would be pretty easy to add a branch for the shared segment of the 20/21/22, which, going by your stats, should add a fair amount of ridership as well.
And as far as RIPTA’s service planning goes, it strikes me as mostly being pretty random.
The streetcar is explicitly a development tool for the Jewelry District. It might even be a net good – one of the local power brokers wants the city to build 7,000 parking spaces in the area instead, and if the streetcar is the alternative to that then its capital cost is actually negative ($126 million for the line minus $140 million for parking).
The tunnel is just ridiculously underused. The biggest difference in ridership between the 42-Hope and 99-North Main is actually from the inner segments – there’s not much ridership from College Hill proper, whereas further out, the 42 actually beats the 99 within Providence proper (the 99 is again busier in Pawtucket). The problem, I think, is that they’re not treating the tunnel well. There’s no schedule posted at the top end of the tunnel, no single route that serves it is all that frequent, and the schedules don’t guarantee consistent trunk headways. Even at the bottom end, they only posted a map after Jef Nickerson drew one. The streetcar is supposed to help fix that, which is why I bring up posting a schedule and a map at the top end as an alternative.
The 99 is actually two bus routes, Pawtucket Avenue and Main Street in Pawtucket. If each were treated (and numbered) as separate routes, the ridership would be far lower than the 42.
It wouldn’t be far lower at all – it would be about the same. The 99 has 5,200 weekday riders, the 42 has 2,500. However, I’ll buy that because of unreliable or inconsistent frequency, College Hill ridership on the 42 is depressed, so the 42 morally should have higher ridership than each 99 branch.
But, as far as streetcar planning goes, the 99 partially parallels the existing mainline, and the 42 doesn’t. Some of the biggest anchors on the 99 and its Main Street branch in Pawtucket are quite close to the rail line, including Mineral Spring, just south of New York Plaza Place, and sort of the Doyle-Olney stretch. Similarly, the 54 has not much ridership originating south of Woonsocket, and much of that ridership is at Lincoln Mall, so Blackstone Valley regional rail with a connecting bus to Lincoln Mall could replace it entirely.
A different kind of frequency and connection or lack of it, are bus line clusters and access from Kennedy Plaza (KP). Thayer St./Wayland Sq, Olneyville Sq and Charles St/Walmart have special schedules for their multiple routes. Thayer and Charles have seven lines each and Olneyville has four. There are a few other destinations that don’t have official coordinated schedules, such as Trinity Sq, Hoyle Sq, and Providence Sta. If you are want to wait for the next bus to Thayer St from KP there are four different lettered stop locations distributed around the plaza that are not close to each other. To decipher which lettered you have to wait at for the next bus, you have to cross reference the schedule and then the KP stop location map. If and when KP gets redesigned the bus route groupings should be clustered together at the same or adjacent KP lettered stop locations.
Routes Destination KP stop and route
51, 52, 53, 54, 55, 56, 72 Charles I_55 / L_51, 54 / O_56 / Q_52, 53, 72
32, 35, 40, 42, 49, 78, 92 Thayer K_42, 49 / M_35, 78 / N_32 / Link_92
17, 18, 27, 28 Olneyville B_17 / E_18 / P_27, 28
11, 20, 21, 22 Trinity Sq A_20, 21, 22 / F_11
13, 17, 18, 31 Hoyle Sq C_13 / B_17 / E_18, 31
50, 55, 56, 57 Providence Sta I_50, 55 / O_56, 57
One concept that might be worth considering for RIPTA service and streetcars: perhaps it makes sense to replace only the busier inner-city parts of bus lines with streetcars, and then, rather than terminating the bus lines from further out at the streetcar terminus and forcing a transfer, the streetcar lines could be extended to the nearest highway so that for local service, you have to transfer, but the bus continues express to Kennedy Plaza. So for example, there would be a streetcar line (or two) to Olneyville Square, where you could transfer to the 17, 19, 27, and 28, which would skip the slow inner-city sections by going on Route 10. Other cases where this could work include routes 51, 54, and 58 in the north, as well as 21 and 22 in the south, and possibly others can be done with some creativity. Since the highway network is there and not going away, you may as well take advantage of it. In fact, looking at the map, this is what it looks like is going on with routes 30 and 31.