Two Sundays ago, instead of going for a bike ride in the nice weather like I should have, I went on a road trip of sorts to look at the condition of the six stations on the MBTA’s Providence Line that don’t have full-height platforms. Those six stations are Hyde Park, Readville (where no trains currently stop), Canton Junction, Sharon, Mansfield, Attleboro, and South Attleboro. I took notes on the site conditions at all six stations so I could make an attempt at guessing the cost of the actual station upgrades required. All six stations are already equipped with “mini-high” platforms and an accessible path of travel to street level and accessible parking via ramps (none have elevators), but the overall condition of the facilities varies greatly. At all stations on the line, signs direct passengers to board trains via the mini-high platforms (which are always at the south end of the platforms) except during inbound peak periods — presumably this is to reduce dwell times by avoiding dropping door traps to open doors at the low platforms. There is also an automated audio announcement system, tied in with the signals, warning passengers of approaching trains. All stations have fencing preventing pedestrians from crossing the tracks at grade. Beyond that, the stations vary greatly; here are my observations.

Hyde Park

The northernmost pre-ADA station on the line is located on Folsom St., south of River St. and a block west of Hyde Park Ave. in the center of Hyde Park. Pedestrian access is served by straight ramps down from the River St. overpass and by side streets connecting Hyde Park Ave. to the station parking lot, which is at rail grade. There is an embankment on the west (southbound) side of the station, which prevents any platform or access expansion on that side. There is a single express track for Amtrak between the northbound and southbound tracks. MBTA buses 32, 33, 40, and 50 serve the nearby business district.

Overall, with the exception of the concrete ramps and the relatively newer mini-high platforms, this station is in poor condition. The low-level platforms lack tactile edging and are cracked and uneven in places. The station signage is worn out, and other finishes and amenities such as lamp standards need replacement. I do not know whether the current ramp system meets MAAB requirements for accessibility, or if it would have to be either replaced or augmented by a second means of egress — especially for the southbound platform, which lacks access to the parking lot. Possibly the station wound need a pedestrian bridge at the south end of the platforms, which would be expensive to construct over the live catenary.

Readville

There are four platforms at Readville, inexplicably numbered 2 through 5. The Providence Line runs past platforms 2 and 3 — platform 4 is for the Fairmount Line and Franklin trains that run via the Dorchester Branch rather than the Southwest Corridor, and platform 5 serves the remaining Franklin Line trains. I say “runs past” because neither Stoughton nor Providence trains currently stop at Readville; any plausible scenario for frequent service on the Providence Line will require shifting most or all Franklin trains to run through onto the Fairmount Line, via the single-track platform 4, because capacity on the three-track Southwest Corridor is limited. In this scenario, especially during the transition period, Franklin Line passengers will want to transfer to the Providence Line trains to access Longwood Medical Area and Back Bay, so improvements should be made to this station to give passengers some weather protection. This is somewhat challenging because the accessible parking at Readville is located between the top of the platform 2/3/5 ramp system and platform 4. Platform 4 is also much too short, just 300 feet, although there is room to extend it further north. The northbound Providence Line platform is at grade with one of the four parking lots serving the station, additional ramps and stairs would be required to connect it to the lot after raising. Finally, in order to maintain the desired all-day schedule on the Fairmount Line without fouling the Franklin Line during peak, an additional platform will be needed for turning Fairmount trains, and it is not at all clear how to make it accessible.

Canton Junction

Like Readville, Canton Junction has a complex system of ramps to provide access to all of the platforms from street level. The station has four platforms, which for some reason are identified with letters A through D rather than numbers; the Stoughton Branch diverges from the main line at the station, and platforms B (Providence northbound) and C (Stoughton southbound) are located in the infield of the junction and accessible only by the ramp system (or the parallel stairs, which are integrated with the structure). The station building is located on platform D (Stoughton northbound), and there are large parking lots both east and west of the tracks. The east lot has a ramp directly to the platform D mini-high, which is just south of the station building, and that ramp would also serve for access to a full-high platform, although stairs would need to be provided north of the building. The only canopy at the station is on platform D, which may account for the surprisingly high passenger loading on the Stoughton Line at this station. The west lot is at grade with platform A, so it would require the construction of additional ramps and stairs, for which there is plenty of room. The mini-highs on all four platforms are farther south and have their own ramps.

Sharon

This was my mother’s station when my folks lived in Sharon and later Foxborough. There are parking lots on both sides of the tracks: one is restricted to Sharon residents and one is available to non-residents. There is a station building adjacent to the northbound track, but set back far enough that it wouldn’t impede platform construction — it would need additional ramps, however, as it is nearly at the north end of the platform and the existing mini-highs are at the south end. The station was renovated relatively recently, but the only wheelchair access between the two platforms involves a very long journey on the Rt. 27 overpass via two ramps; I’m not sure this really meets MAAB standards, but if it does, that would make this station relatively simple to upgrade. (Moving wheelchair users to the north end of the platform would significantly improve their travel experience when parked in the lot opposite their desired platform, but it really feels like disabled passengers were treated as an imposition in the design of this station.) Both platforms are at grade with the parking lots and would need additional ramps and stairs.

Mansfield

Oy. What can I say about Mansfield? The MBTA is nearly finished spending $7.1 million to renovate this station and didn’t build proper high platforms. This has been an ADA and MAAB requirement since the 1990s, yet somehow they managed to weasel out of it in 2015. The station does have mini-highs on both tracks, and it has a new system of ramps and stairs to connect the MBTA parking lot (south and west of both platforms) to the station; there is also a large private parking lot adjacent to the southbound platform. This site is highly constrained, with a viaduct over Chauncy St. at the south end of the platforms and a single-track junction with the Framingham Secondary immediately north of the southbound platform. This station has been widely reported as being a problem for clearances for military freight trains (heading from Framingham to Cape Cod), but if the hinged platform edges on the brand new mini-highs are sufficient for STRACNET purposes, there seems no reason the same treatment wouldn’t be sufficient on full-length platforms as well — especially since the Framingham Secondary can only access the southbound track. Longer term, it would be preferable to build a third track where the current southbound platform is: the right of way broadens to four tracks just north of the station, and giving intercity trains a bypass would not be a bad thing, but this is likely too expensive a project for the short term, so conflicts with Amtrak will have to be managed through careful scheduling.

On the positive side, there seems to be a good amount of transit-oriented development taking place on the parcels adjoining the station, and it sees a substantial number of daily boardings, so building full high platforms would be entirely justified on the basis of the current service pattern, never mind a modern service. If only the state hadn’t just blown millions building the wrong thing! (The people to blame for that error of judgment, by the way, are Brian Shortsleeve and Stephanie Pollack.)

Attleboro

The line runs through Attleboro on a four-track embankment; the original station structures on both the northbound and southbound tracks still stand, but the southbound platform has been relocated farther south and the original station (now leased out to office tenants) is no longer accessible from the platform. The two station buildings are bracketed on either side by underpasses for city streets, Mill St. on the north and South Main St. on the south, and the northbound platform continues across Mill St. for some distance. There is a substantial parking lot on the southbound side, and GATRA operates a substantial bus station adjacent to the lot, which they also own. A smaller lot serves the northbound platform and station building, but there is no direct access between the two platforms; passengers must descend to South Main and cross the tracks via the underpass.

South Attleboro

The sixth and final station is in many ways the worst. South Attleboro is located hard up against an embankment to the south (constricting the northbound platform) and in the shadow of Newport Ave. (Rt. 1A) overhead just to the east. There is an interchange on divided Rt. 1A just to the north, and the road past the station (at grade with the southbound platform) is busy with traffic from I-95 south to a power center on the northbound side of 1A. The parking lot for South Attleboro station is across this street, and there is a single crosswalk connecting it to the southbound platform; the beg button was not functioning when I visited, but perhaps the signal only operates on weekdays.

The only way to access the northbound platform is via a single structurally deficient pedestrian overpass and ramp system that connects to the parking lot but not to the southbound platform — the stairs to the southbound platform have rusted out and are blocked off at both ends. The stairs on the northbound platform have also rusted out, but at least the ramp down from the overpass connects directly to the mini-high at the west (railroad south) end of the platform. I cannot believe that this station meets code, and it will require a substantial investment whenever any change is made, to add a second means of egress from the northbound platform and likely a second overpass. I’m somewhat surprised that it’s even still open at all — and yet, it serves more than a thousand daily boardings, so clearly there is substantial demand. (A license-plate survey would be interesting: are these passengers mostly from Rhode Island, in which case perhaps they would prefer to go to the new Pawtucket station when it opens?)

Upgrade costs

I’m no construction estimator, nor am I an architect or civil engineer. But I can make a semi-educated guess about how much high platforms ought to cost. My most basic assumption is that a train platform can be effectively constructed like a reinforced concrete box girder, but with integrated overhangs for the foundation slab and the platform surface itself. This allows for a variety of construction techniques, both cast-in-place and pre-cast, and the hollow cross-section reduces the materials cost and provides a convenient chase in which to run cabling for platform lighting and signage. Segmental construction makes it possible to keep stations open during construction, if necessary. I estimated that the materials cost of two such platforms would be about $350,000 (at $200 per cubic yard of concrete), and figured that labor would be roughly twice that (depending on construction sequencing and whether it’s cast on site or precast). When you add in new platform signage, canopies, precast stairs and ramps, safety railings, and so on, a figure of $2 million would seem to be reasonable. Why, then, do MBTA station projects seem to cost so, so much more? I don’t know, and I would like to see a more detailed analysis of what is actually going into these projects.

That said, I can at least make a stab at the total upgrade costs for the whole line.

Hyde Park gets the fewest daily passengers of any of the stations under discussion, so it would seem like an obvious candidate to close for a few weeks in the summer and just build everything in place. However, it will be somewhat more challenging to construct due to the lack of easy access to the southbound platform for construction equipment, and the need to maintain clearance for through trains. Doubling my base estimate to give an upgrade cost of $4 million here.

Readville should be the easiest, because it doesn’t have scheduled service at the moment, but it would be silly to upgrade the Providence/Stoughton platforms and not upgrade the Franklin/Fairmount platforms as well. There are similar site constraints to Hyde Park and additional problems with the Franklin/Fairmount lines layout, so I’ll go for $8 million — add another $20 million if you want to augment the ramp system with elevators here. I’d probably also add at least $10 million to twin the Franklin Line overpass, which is probably necessary in the long term to support frequent service on the Fairmount (which needs to have a second platform added when the remaining Franklin trains get through-routed).

Route 128 obviously needs no work; it’s already a high-platform station. Canton Junction should be relatively easy, even though there are four platforms, because all of the platforms are accessible without crossing the Corridor and its overhead electrification. So I’m going with a flat $4 million, on the assumption that the existing ramp system is adequate and can be easily altered to support the new platform height. Sharon and Mansfield should be similarly easy, although Sharon may require an additional investment in vertical connectivity. Assume there’s an extra $1m in costs at Mansfield related to the hinged platform edge system, and you’re at $5m for the two.

Attleboro should be relatively easy, and there’s already adequate egress, but both platforms have sections that may be difficult to access from the street. This may be an ideal opportunity to do cast-on-site and then close the line and shut off power to crane a new platform in in large pieces. Attleboro is the third-busiest station on the line so obviously closures need to be kept to a minimum. Call it $4m.

Then there’s South Attleboro. It has all of the same problems as Hyde Park plus the decrepit overpass needs substantial refurbishment. The street that passes between the southbound platform and the parking lot will need to be narrowed to widen the platform and provide a safe ramp down to street level. If it cost less than $8m I would be very surprised.

That brings the total to $43 million. Rail Vision is probably estimating it at an order of magnitude higher, which is MassDOT’s way of ensuring that either nothing gets done, or all their friends in the contractor and consultant industry make a hefty profit. Add to that the cost of upgrading Sharon substation (maybe $20m?), building a new maintenance facility for EMUs (might as well do it in Pawtucket, since Readville will be otherwise occupied for a while), perhaps $30m, and actually acquiring the new EMUs (24 120-meter EMUs, including the Rhode Island shuttle and two spares), $288m. All told, $360 million for a complete modernization of the Providence Line. These investments will also support electrification of the Fairmount Line as the second step in the full electrification of the network.