Tuesday, July 31, 2018

Bus yards vs TOD: where is the best place to store buses?

Boston has a bus problem. Beyond narrow, congested roads and routes which traverse several jurisdictions—in some cases half a dozen in the span of a single mile—there are simply not enough buses to go around. At rush hour, some MBTA bus routes only have service every 20 to 30 minutes, despite crush-capacity loads on the vehicles serving them. To add significantly more service would require the MBTA to add additional buses to the fleet, but procurement of new vehicles is not the rate-limiting factor. The larger issue is that the MBTA’s bus storage facilities are undersized and oversubscribed, so adding new buses would require adding additional storage capacity to the system, a high marginal capital cost for any increase in service.

Before doing this, the MBTA may be able to squeeze some marginal efficiency from the system. All-door boarding would reduce dwell times, speeding buses along the routes. Cities and towns are working with the agency to add queue jumps, bus lanes and signal priority, steps which will allow the current fleet to make more trips over the course of the day. Running more overnight service would mean that some number of buses would be on the road at all times of the day and night, reducing the need to store those buses during those times (although they might need to be serviced during peak hours, and may not be available for peak service). Still, all of this amounts to nibbling around the edges. Improving bus service may result in increased patronage, and any additional capacity wrung out of the system could easily be overrun by new passengers. The MBTA’s bus system is, in essence, a zero-sum game: to add any significant capacity, the system has to move resources from one route to another: to rob Peter to pay Paul.

Furthermore, Boston’s bus garages are antiquated. In the Twin Cities—a cold-weather city where a similarly-sized bus fleet provides half as many trips as Boston (although about the same number of passenger miles)—nearly every bus garage is fully-enclosed, so buses don’t sit outside during cold snaps and blizzards as they do in Boston. Every facility there has been built since 1980, while several of the MBTA’s bus yards date to the 1930s; some were originally built for streetcars. Boston desperately needs expanded bus facilities, but it also needs new bus garages: the facilities in Lynn, Fellsway and Quincy are in poor condition, and the Arborway yard is a temporary facility with very little enclosed area.

However, what Boston’s bus yards lack in size or youth they make up for in location. The MBTA bus system is unique in the country in that there is no bus service through downtown: nearly every trip to the city requires a transfer from a surface line to a rapid transit line. In the past, elaborate transfer stations were built to facilitate these transfers, with streetcar and bus ramps above and below street level (a few vestiges of this system are still in use, most notably the bus tunnel at Harvard), with bus routes radiating out from these transfer stations. When the Boston Elevated Railway, the predecessor to the MBTA, needed to build a streetcar yard, they generally built it adjacent to a transfer station, and thus adjacent to as many bus routes as possible. Many of these have become today’s bus yards, and the MBTA has some of the lowest deadhead (out of revenue service) mileage to and from the starts of its routes.

From a purely operational standpoint, this makes sense: the buses are stored close to where they are needed. But from an economic standpoint, it means that the T’s buses occupy prime real estate. Unlike rail yards, which need to be located adjacent to the lines they serve, bus yards can be located further away. While this introduces increased deadhead costs to get the buses from the yard to the route, it frees up valuable land for different uses. In recent decades, the T has sold off some of its bus garages, most notably the Bartlett Yard near Dudley and the Bennett Yard near Harvard Square, which now houses the Kennedy School. The downside is that the T currently has no spare capacity at its current yards, and needs to rebuild or replace its oldest facilities.

While the agency has no concrete plans, current ideas circulate around using park-and-ride lots adjacent to rail stations for bus storage, including at sites adjacent to the Riverside and Wellington stations. The agency owns these parcels, and the parking can easily be accommodated in a nearby garage. The issue: these parcels are prime real estate for transit oriented development, and putting bus garages next to transit stations is not the best use of the land. Riverside has plans in place, and Wellington’s parking lot sits across Station Landing, which has hundreds of transit-accessible apartments.

In addition to what is, in a sense, a housing problem for buses, the Boston area has an acute housing problem for people. The region’s largest bus yards are adjacent to Forest Hills, Broadway and Sullivan Square: three transit stations with easy downtown connections. These issues are not unrelated: there are few large parcels available for housing or transit storage (or, really, for any other use). If the region devotes land to housing, it may not have the ability to accommodate the transit vehicles needed to serve the housing (without devolving the region in to further gridlock). If it uses transit-accessible land for storing buses, it gives up land which could be used for dense, transit-accessible housing. What the transit agency needs are sites suitable for building bus depots, on publicly-owned land, and which would not otherwise have a high-level use for housing.

Consider a bus maintenance facility: it is really something no one wants in their back yard. And unlike normal NIMBYism, there actually some good reasons for this: bus yards are noisy, have light pollution, and operate at all times of day, but are especially busy for early morning operations. An optimal site for a bus yard would be away from residences, near highways (so the buses can quickly get to their routes), preferably near the outer ends of many routes, and not on land which could otherwise be used for transit-oriented development. It would also avoid greenfield sites, and preferably avoid sites which are very near sea level, although if necessary buses can be stored elsewhere during predicted seawater flood events.

The MBTA is in luck. An accident of history may provide Boston with several locations desirable for bus garages, and little else. While most sites near highways don’t have enough space for bus yards, when the regional highway system was canceled in the early 1970s, several interchanges had been partially constructed, but were no longer needed. While portions of the neighborhoods cleared for highways have been, or could be, repurposed in to developable land, the “infields” of highway ramps is not generally ripe for development. Yet they’re owned by the state, currently unused, convenient to highways and unlikely to be used for any other purpose. For many bus routes, moving to these locations would have a minimal effect on operation costs—deadhead pull-in and pull-out time—and the land will otherwise go unused. Land near transit stations is valuable. Land near highways is not.

Building bus yards in these locations would allow the T to add vehicles to the fleet while potentially closing some of its oldest, least-efficient bus yards, replacing them with modern facilities. They wouldn't serve all routes, since many routes would still be optimally served by closer-in yards with shorter deadhead movements to get the buses to the start of the route. (To take this to an extreme: it would be very cheap to build a bus yard at, say, the former Fort Devens site, but any savings would be gobbled up by increased overhead getting the buses 35 miles to Boston.) Highway ramps are optimal because it allows buses to quickly access the start and end of routes, many of which, by history and happenstance, are near the highways anyway.

Most importantly: moving buses to these locations would enhance opportunities for additional housing, not preclude it. Building thousands of new housing units adjacent to transit stations pays dividends several times over. It increases local tax revenues and also creates new, fare-paying transit riders without the need to build any new transit infrastructure. Finally, by allowing more people to use transit for their commutes, it reduces the growth of congestion, allowing people driving—and people riding transit—to move more efficiently.

Specifically, there are five highway sites in the region which could be repurposed for bus fleet facilities:

  • Quincy, in between the legs of the Braintree Split
  • Canton, on the aborted ramps of the Southwest Expressway
  • Weston, where the new all-electric tolling has allowed for streamlined land use
  • Burlington, in the land originally planned for the Route 3 cloverleaf
  • Revere, in the circle where the Northeast Expressway was originally planned to branch off of Route 1 through the Rumney Marshes.

In more detail, with buses counts from the MBTA’s 2014 Blue Book. These are in-service buses required, so the total number of buses at each location, accounting for spares, would be 15 to 20 percent higher. The system currently maintains approximately 1000 buses.

Quincy (67 buses)

All 200-series Quincy Routes

The current Quincy garage serves the 200-series routes, with a peak demand for 67 vehicles. The current garage is in need of replacement. The current yard takes up 120,000 square feet on Hancock Street, half a mile from Quincy Center station. This could easily be accommodated within or adjacent to the Braintree Split, with minimal changes to pull-out routes. Serving additional routes would be difficult, since the nearest routes run out of Ashmont, and pull-out buses would encounter rush hour traffic, creating a longer trip than from the current Cabot yard.

Canton (35 buses)

Routes 24, 32, 33, 34, 34E, 35, 36, 37, 40

This would be a smaller yard and would probably only operate during weekdays with minimal heavy maintenance facilities, but would reduce the overall number of buses requiring storage elsewhere.

Weston (71 buses)

Routes 52, 57, 59, 60, 64, 70/70A, all 500-series express bus routes.

With the recent conversion to all-electronic tolling on the Turnpike and different ramp layout, the land is newly-freed, plentiful, and many buses serving this area have long pull-out routes from Boston. The portion between the two branches of the Turnpike and east of the 128-to-Turnpike ramp is 500,000 square feet, the same size as the Arborway Yard, and there's additional room within the rest of the interchange. Without a bus yard west of Boston, any route extending west or northwest would benefit from this yard.

Burlington (50 buses)

Routes 62, 67, 76, 77, 78, 79, 134, 350, 351, 352, 354

These routes utilize serve the northwest suburbs, but most are served by the Charleston and Bennett divisions in Somerville. Most routes would have significantly shorter pull-outs.

Revere (157 buses)

The two oldest bus garages north of Boston are Lynn and Fellsway, which account for a total of 125 buses and about 200,000 square feet. They are both centrally-located to the bus network, so moving buses to the 128 corridor would result in longer pull-outs, except for a few routes noted above. However, the circle where Route 1 turns northeast and the Northeast Expressway was originally planned and graded towards Lynn across Rumney Marshes has 750,000 square feet, and the extension towards the marshes more. The fill is far enough above sea level to not worry about flooding, and grade separation allows easy exit and entry on to Route 1. Some buses may make sense to base at the Route 3 site, particularly the 130-series buses. In addition to the Lynn and Fellsway buses, this site could take over for many routes currently operating out of the Charlestown yard, freeing up capacity there for other uses.

Other routes served by the Charlestown yards would face somewhat longer pull-out times from Revere, but given the development potential in Sullivan Square, the T could consider downsizing the yard facility there and moving operations to a less valuable site. This site, at more than one million square feet, could likely replace the Charlestown bus facility entirely.


  1. The proposed Wellington bus facility is hardly impactful of Station Landing, and is not itself a TOD candidate. The yard would be 1000 ft. away with the expanded Orange Line yard in the middle, pinned in along the river well behind the Wellington Station busway and replacement vertical garage. Layout shivved back in that corner amid all those other facilities makes it disconnected and extremely low-value land for building anything mixed-use...so this backlot is exactly the kind of misfit large parcel lacking appropriate commercial options where you would WANT land use devoted to a bus yard. For 190+ buses and enabling 60-foot bus capability on any current Charlestown route that's a hell of a good deal, and substantially bigger than the Revere proposal. Not to mention it outright closes Fellsway and downsizes Lynn to day storage to-boot, accomplishing consolidation of other facilities MUCH more impactful to their surroundings and, in Fellsway's case, allowing immediate sale of an above-average located redev parcel.

    And why Weston when Watertown garage is already available and can have its non- location-specific machine shop functions relocated literally anywhere to make room for 90 buses? Watertown's not a good TOD site either because of the substantial grade difference between the Water St. and Nonantum Rd. sides of the parcel. The current garage building abuts against the Nonantum curb like it does because the grade preventing sidewalk or curb cuts on that side of the road make it useless for any street-facing development. Despite superficially convenient appearance, it's a thoroughly misfit parcel that will have a hard time attracting commercial developers. When that yard can be rearranged to stuff bus parking further away from Water St. for less-impactful noise/fumes AND possibly absorb all trackless trolley fleet heavy-repair to downsize or eliminate North Cambridge Garage, why reach FIRST for Weston? It's further west than any MBTA bus roams and is situated in a staunchly anti-development bedroom community that's going to come nukes-loaded at the first mention of such a facility.

    South region I completely agree needs some work and realignment because Quincy's a tough case, but north and west regions don't call for a big pensive thinkpiece. Devil's advocacy isn't going to make TOD panaceas out of intrinsically very awkward development parcels. When those very parcels are already hosting very centrally-located T facilities...well, where exactly is bigness of the dilemma with the Watertown or (especially) Wellington sites?

  2. How do you increase bus capacity while using the same storage space?

    Add a second floor to your buses.

  3. Fline, I'm not sure what to make of this …

    The Wellington parking lot is "extremely low value" land? What makes it any lower value than the development on the other side of the yard? Or than the casino site? It's on the water. It's next to transit. It has good highway access. This is just ignorant.

    As for Watertown, same deal. That parcel is adjacent a walkable town center with good transit connections, as well as being across the street (a relatively narrow DCR parkway) from the river and parkland. Yet that's somehow undesirable?

    The Weston site is in the middle of a highway. Weston can complain all they want, but it would be entirely on state land, so there's not much they can do.

    1. That's a spurious comparison. Station Landing has a complete three-block street grid with 2 egresses to/from Route 16 East, 2 egresses to/from Route 28 north, and 1 complete signalized egress to both directions of 28. Wellington's current configuration only affords 1 two-way egress to the single access road, and 1 one-way ingress. The Wellington acreage is significantly lower value because it cannot possibly be reconfigured to allow for equivalent development density as Station Landing with only a maximum of 2 egresses to one single geometrically constrained access road that gets solidly jammed during rush. Assuming that any traded-in development acreage needs to send the station parking vertical, the footprint of the new garage will ensure that any access to the rear development will need to share driveways with the garage making getting in/out of the development a royal pain at rush. Station Landing, having multiple egresses, almost always has a route for escaping the worst of Wellington Circle traffic with the President's Landing signal on Fellsway being the fail-safe.

      That is a hell of a big difference, Ari. A hell of a big difference not shared by any of those other area sites you name like the casino. Having significantly, unresolvably poorer access than all other area TOD makes the Wellington back lot a very significantly lower value property for developers and harder sell for TOD than all other area sites with superior access. It is not "ignorant" to point out how they are not the same; the name-calling is not warranted.

      Watertown I already explained has complications with the significant grade difference between the Water St. and Nonantum Rd. sides of the parcel. It would be difficult to mass up against Nonantum to provide any sidewalk or curb-cut access there...the hillside, of course, being why there is no sidewalk whatsoever on that entire side of Nonantum. That limits potential development to only what can face Water St. and sacrifices most upside of closeness to the river and parkland. Does it totally nullify the utility of the parcel? No. Does it even lower the value to the same degree as Wellington's much more constrained single access point? No...Water St. does run the whole length and affords multiple outlets. But it does induce complications at subdividing the slab when Nonantum can't take facing development. So long as Colonial Buick and a few other crud auto parcels are still available on the Galen-Water block, prospective TOD land in the neighborhood is not yet at a scarcity threshold where Watertown Yard is a high-leverage redev parcel. The MBTA's timetable for finding bus storage is likely shorter-term than EITHER the cresting of Watertown Yard's TOD land value OR any MassHighway plans for compacting the Weston interchange to serve up new usable land out there (and figure out means of access to that land that avoids any use of town-control roads where the Westonites have attack vectors for blocking or slowing the build). There is very much a sliding scale of necessities in-play there re: land use value worth basic acknowledging. While you can make a reasoned argument of TOD value cresting first and taking priority, it's a hell of a lot more dynamic and less absolutist than "same deal".

    2. Weston already got reconfigured, the ramp from 90 westbound to 95 got moved from the near side of the 95 to 90-EB overpass to the far side of Park Rd, to provide more merging distance. Given the bridges over the Charles aren't likely to be moved, the land in that "infield" can now be put to better use. In fact, just the land from the old ramp to 95 would be enough for a decent sized bus garage with potential for ramps directly to/from the Pike, that wouldn't even get in the way of a realignment of the Turnpike Extension to straighten it out some through that area.

    3. I have no idea what you are smoking, but saying that development won't occur next to Wellington because traffic is … well you told me not to use that word.

  4. Given the rapid development of fully-electric transit buses, bus maintenance facilities of the future will be far less polluting (both greenhouse gas and noise) than the noisy, smelly facilities of the present. The MBTA should not be in any hurry to sell off valuable parcels but instead should look to redevelop these sites to include unobtrusive maintenance and storage capacity for electric buses. Done right, these facilities should be no noisier than a parking garage.

  5. Is there the ability to add bus garages to the proposed new TOD developments near end of lines? Facilities for new electric buses could possibly be incorporated into the design. Developers may even donate a bus as part of an impact assessment or for bonus density or other incentives. The fleet would be dispersed but it could be dispersed in a manner that the buses are along the lines they serve to maintain low "dead head" periods.

  6. Wow very nice and Very informative post, thanks for sharing i liked your blog

  7. Ya know, I used to work for the Mass. Highway Department and I know for a fact that the park 'n' ride lot next to the MassHighway maintenance yard could accommodate more than 67 busses including repair facilities that the land within the Braintree split of Route 93 at Routes 3 and 128. Another advantage is that whereas taking land within the split would pose a negative environmental impact (loss of trees, nighttime light distraction of motorists, and who wants to look at busses?) and a bus operations nightmare (merging into 70 mph traffic or sitting in 5 mph traffic). Using the MassHighway yard car park OTOH would only be an improvement.