» A new station on Boston’s Orange Line prepares for opening, but infill stations of its type are all too rare.
Want to know a secret? One of the best ways to increase transit ridership at a reasonable price requires little additional service. It requires no new line extensions. And it can be done to maximize the value of existing urban neighborhoods.
This magic solution comes in the form of the infill station–a new stop constructed along an existing line, between two existing stations. Next week, Boston’s MBTA transit agency plans to open a new stop, Assembly Station, along the Orange Line in Somerville, a dense inner-ring suburb just to the northwest of downtown Boston.
Assembly is the latest in a series of recent infill stations in the U.S. located along older heavy rail lines whose other stations were generally constructed decades ago. Washington, D.C.’s NoMa Metro Station opened in 2004; the San Francisco region’s West Dublin/Pleasanton BART Station followed in 2011. In Boston, new stations have been constructed along the upgraded commuter rail-becoming-regional rail Fairmount Corridor. And Chicago has had success with the opening of two infill stations in 2012, the Morgan Station in the city’s West Loop and the Oakton-Skokie Station in the northern suburbs.
Yet those expansions are exceptions to the rule. Two infill stations are currently planned in Northern Virginia, at Potomac Yard along the Metro in Alexandria and at Potomac Shores along the VRE commuter line, and one new station is under construction along the Green Line in Chicago.
But few other cities or transit systems are even considering the possibility of investing in infill stops, even as line extensions are proliferating around the country. That’s a big disappointment.
The advantages of infill stations result from the fact that people are simply more likely to use transit when they’re closer to it — and from the fact that the older transit systems in many cities have widely spaced stations that are underserving potentially significant markets. Erick Guerra and Robert Cervero, affiliated with the University of California-Berkeley, have demonstrated that people living or working within a quarter mile of a transit station produce about twice as many transit rides as people living or working more than half a mile away. In other words, with fewer stations on a line, the number of people willing to use public transportation as a whole is likely reduced.
Assembly Station, which has been in the works for several years, promises significant benefits — 5,000 future daily riders taking advantage of a 10-minute ride to the region’s central business district, at a construction cost of about $30 million. The station fits in the 1.3-mile gap between two existing stations and is the first new stop built along Boston’s T rapid transit network in 26 years. When combined with the $1.7 billion Green Line light rail extension planned for opening later this decade, 85 percent of Somerville’s residents will live within walking distance of rapid transit, up from just 15 percent today.
The cost-per-rider comparison between the two Somerville projects is indicative of the value offered by infill stations: While Assembly Station cost about $6,000 per rider served, the Green Line Extension will cost $38,000 per rider served — six times more. Both projects will provide benefits, but the cost-effectiveness of infill stations in terms of attracting riders is clear. While infill stations will reduce transit speeds to some extent, within reason the number of new riders they attract will more than make up for the change.
Assembly Station was made possible in part thanks to a $15 million contribution from Federal Realty Investment Trust, which is building a $1.5 billion mixed-use community adjacent to the station. This Assembly Row project will eventually include 2,100 housing units, 500,000 square feet of retail, and 1.75 million square feet of office space, in effect creating a transit-oriented mini-city in an area that was once home to an automobile plant and, after that, a strip mall.
The ingenious decision to combine the creation of a dense new development with a new transit station encourages the production of a virtuous cycle of more people living near transit who thus are more likely to use transit.
There are many other places throughout the country where there are similar opportunities for new infill stations. San Francisco’s BART studied a new subway station at 30th Street and the Mission years ago but has done little to act on the idea. Other cities have the physical conditions that are right for infill stations but little momentum to implement them. Portland’s light rail system, for example, has several long inter-station gaps: The distance between Lloyd Center and Hollywood/NE 42nd (east of downtown) is 1.7 miles, certainly long enough to justify a new station in between. In Atlanta, similarly, the gap between Arts Center and Lindbergh Center on the Red and Gold Marta lines is 2.7 miles, passing through a zone of potential redevelopment.
These examples are just the tip of the iceberg; there are dozens of similar situations around the country.
Transit agencies looking for ways to maximize the use of their existing lines should look to literally fill the gaps between their existing stations. In doing so, they offer the opportunity to build additional ridership and spur redevelopment.
Image at top from MBTA.
35 replies on “With infill stations, older transit agencies extend their reach”
“Cost-per-rider” is meaningless, unless you mean the total riders the project will get over the entire lifespan of the project (which could be 100 years). We should use either “cost-per-rides/day” or “cost-per-rides/year”.
Cost per rider is generally short for cost per typical weekday rider at some future date. From the link Assembly is projected to serve 5k riders per typical weekday by 2030. At 29.2 million that’s about 6k per typical weekday rider.
So people should say cost/weekday rider then :-)
The trouble with using is that people then compare that with the fare paid per rider, and start foaming at the mouth….
Not mentioned in the article: this disbenefit to existing users. Adding stops will increase travel time, and hecne reduce ridership (slightly) at other stations. This is particularly crucial when the travel time is close to an alternative mode (typically car travel).
Yeah — this means there are only a few places where this sort of infill makes sense.
There are several good places for infill stations in Chicago: mostly places on the L where former stations were ripped out, which have large, dense populations. The L trains don’t go that fast anyway and the stop is well worth it. (A few of these have been built already.)
Most other cities don’t have similarly obvious infill opportunities.
In DC, there has been some conversation and study of a few additional infill stations, notably:
Oklahoma Avenue: would be on Orange/Blue/Silver line between the existing Stadium/Armory Station and the Anacostia River; station would provide better access for the northeastern parts of Capitol Hill/Trinidad, and perhaps whatever development eventually replaces the RFK Stadium parking lots.
Grant Circle: would be on the Green/Yellow at New Hampshire Ave and Illinois Ave; would provide second station in Petworth neighborhood of NW DC.
Jefferson Memorial/East Potomac Park: would be on the Yellow line near where it emerges from its tunnel and crosses the Potomac; station would provide enhanced access to Jefferson and FDR Memorials as well as East Potomac Park; could reduce crush loads at Smithsonian station during major Mall events (e.g., Cherry Blossom festival; several runs/races).
As far as I know, none of these are in very serious planning, though at least some were identified on the District’s recent “MoveDC” transportation vision plan.
The big things the two DC infill stations (NoMa and Potomac Yard) have in common are: 1) above-ground track, simplifying construction and lowering construction costs, 2) huge opportunities to upzone and build densely around the new station, and 3) a financing mechanism for the station that ties the development rights to the project execution, meaning that the surrounding land owners are invested in maximizing the value from that land.
Alas, I’m not sure that many of the other infill station proposals hit on those elements. An infill station near Grant Circle would be much harder to build (due to the need to build around an active subway tunnel) and the residential character of the surrounding neighborhood, while decently dense, is not a likely candidate for large upzoning in the current comprehensive plan.
Based on those criteria, the Oklahoma Ave station idea in the RFK parking lot has a lot of promise, but that depends on the feasibility and legal wrangling to develop the RFK site. Further to the east, the old Benning Road Pepco plant will eventually be decommissioned and will need clean-up, but that would also make a good option for a River Terrace infill station (potentially as a transfer station in the future if the Separated Blue Line proposals advance).
The WMATA Technical Advisory Group (TAG) included possible future in-fill stations in their studies several years ago. The 3 that showed potential for acceptable ridership were:
1. Kansas Ave on the Red Line based on 2040 projections for land use around the station. The Red Line is above ground there so it could be done without breaking the bank.
2. St. Elizabeth’s Hospital but this was assuming the DHS complex is built at St. Elizabeth and Congress is stalling funding on construction for consolidating DHS there.
3. Oklahoma/Benning but only “if the station is well accessible from Benning Road”.
Kansas Ave and Oklahoma/Benning may be the most viable locations for future in-fill stations after Potomac Yards, but, yes, there are no serious planning or feasibility studies for either location that I am aware of.
The cost of New York Avenue, also above ground, was higher than the cost of an underground infill station in Seoul.
I wouldn’t add a MAX station between Lloyd Center and Hollywood, for the following reasons:
1) The area already has excellent bus service, including parallel bus service that connects to both MAX in both directions.
2) This is part of the main MAX trunk line, and there’s enough “MAX is slow downtown” complaints.
3) Adjoining land uses are part of a historic district, limiting upzoning potential.
4) As the MAX lines are sandwidched between a freeway and a freight rail line, putting a station anywhere in the stretch would be very expensive.
If anything, the MAX Blue/Red Line should see some stations removed, particularly downtown. (Buh-bye, Kings Hill/Salmon Street).
And of course–#2 is the main tradeoff when considering this sort of thing: Adding stations slows down the service. If it makes new customers and destinations available, it may be worth it. If it’s for the convenience of relatively few, but the inconvenience of many more, perhaps not.
OTOH, I think a Red Line station near 82nd and Airport Way would make sense, and there are few other places along the MAX system where an additional stop would be useful. (TriMet has added stops along existing lines, and the Orange Line opening next year includes one planned “future” stop, to be installed if and when sufficient development around the station occurs).
One consideration for the Boston system is that the MBTA, especially the Orange and Red lines, and to some extent the Green Lines, are already running at or over capacity. Adding riders is not necessarily what the MBTA needs. In fact, expensive risks might be a better bet for the viability of the T in the long term.
Yeah, I was going to say this. This is in effect the same problem as the Green Line — adding riders to a line that is already overloaded may not actually do much overall good. Money would be better spent on increasing capacity (the money budgeted to purchase new Orange Line cars will just replace old cars and not add capacity to the system).
The benefit to the community is that developers in Somerville will make money and the City will expand its tax base, but that money is not going to go back into the MBTA, and if it hurts service on the entire line, will that limit dense development elsewhere (I’m not sure)?
I also doubt there are a lot of places where infill stations could be placed. Lines aren’t often built with 1.3 mile gaps between stops.
The MBTA Red Line’s problem is the antiquated signalling system that allows trains to run no closer than 9 minutes apart. Replace the signalling, and you can quadruple the frequency and capacity without any further costs.
Similarly, Orange Line headways are every 6 minutes, and could easily be 2 or 3 minutes.
9 minutes per branch, 4.5 minutes on the trunk.
The ridership estimate on the Green Line Extension here is probably lowballed. I’d like to check Yonah’s source here. This route is (partially) replacing the streetcar routes which Cambridge and Somerville were *built* around; due to the already-existing supportive land usage, it’ll get extraordinarily high ridership immediately.
I suspect that the quote is “per new rider”, because the people served by the Green Line Extension will largely be switching from buses.
Assembly Square station is great, but 3/4 of its catchment area is dead (there’s a river there). It’ll underperform relative to the estimates. More likely, it’ll cannibalize ridership from Sullivan.
Regarding overcrowding, the Green Line extension is viable, because it’s using capacity in the “anti-peak direction”, by extending to the far side of downtown Boston. The only other line which could be extended in this manner in Boston is the Blue Line.
A BART infill station near 30th + Mission looks great on a map but strikes out as follows :
1) BART’s own models show that the station would be a schedule breaker;
2) The tunnel in that area is sloping and would require extensive rebuilding to get the platforms sufficiently level; and
3) the local NIMBY’s are NOT normal ones – these are more like monsters raised on nuclear steroids.
The original construction in San Francisco, especially in the Mission District, left an extremely sour taste in people’s mouths. The area has few choices for the necessary work spaces that a tunnel rebuild would require. Also, the car and truck lobbies would be up in arms due to a major freeway ramp (I-280 feeder) nearby that would get hammered by construction activity. I lived near there (21st Street) and can see the symmetry of such a station.
BART has two greater needs in S.F. – the chipmunk cheek platforms at Montgomery + Embarcadero Stations and the turnback tracks that should have been put under the Van Ness Stn.
Not to mention the several hundred million dollars (!!) that this would cost versus just a few million for these other infill examples. BART dismissed this idea for good reason.
The usual way for digging station boxes requires huge holes on the surface and many truckloads of dirt being hauled away to either a dump or a processing facility to produce gravel and fines. Has anybody tried to do a sort of keyhole surgery where the dirt gets hauled away on the subway’s own tracks ?
I suspect that it might be more expensive to go underground but it would have the benefit of less surface traffic. This could then be extended to include a mobile batch plant for concrete production and crew shuttles. My reason for asking the above question has to do with building the long-needed turnback tracks under the Van Ness Stn. here in S.F. The usual surface approach would be a mess but a sub-surface approach would be out-of-sight+out-of-mind.
Sure, this is possible. This is known as a “cavern” and it often requires complex engineering to hold up the roof while you dig. Much easier in rock than in soft soils. All things equal, it offers no cost savings over a cut-and-cover process. You only do it if you need to build a station very deep, or if the stuff on the surface cannot be torn up.
This is how New York’s East Side Access is being dug; spoil is removed via conveyors all the way down the tunnel and under the East River to Sunnyside in Queens where it gets hauled away.
It’s theoretically possible to build a station using tunnel boring machines, but it requires building one extra-large tunnel (with room for both track and platform), which means again, it’s not usually cheaper than the standard way.
“Conventional” mining of tunnels is slow and expensive, and will *never* be cheaper than cut-and-cover. TBMs are much cheaper and faster. However, the practicality of such large tunnels seems to be a problem so far.
The reason I asked my question is because there are places where infill facilities, stations or turnback tracks, are needed but a cut-and-cover approach is ruled out due to surface conditions (densely built city, wet weather, fragile park, etc.).
In the mining industry there have been new entrances built using a drift (aka side tunnel) from below and a drilled diagonal shaft from above. Once the drift and the narrow shaft are connected then the shaft can be expanded with the spoil falling down to the drift for removal via the mine’s tunnels to the waste handling facility.
Building turnback tracks for BART in S.F. would be non-starter using the usual surface approach. But sneaking up on the goal, while it would be time consuming and costly, would avoid the political firestorm that a surface approach would trigger. The side platforms that BART would like to add to the Montgomery + Embarcadero Stns. would also benefit from the underneath approach.
P.S. It’s ironic that one of the busiest stations in the BART network, Embarcadero, was an extra that San Francisco had to pay for. The original plans didn’t have a station in that location. In a way, that station is the ultimate in infill stations.
You’re talking like hauling/carting out spoil on special trailers on subway loading gauge? Certainly doable but you’d need special equipment manufactured to run in the tunnels on the electrical voltage, gauge, etc. And then be very very careful not to spill any spoil on the tracks and electrical supply and signals.
I’d imagine that this is how London built some of their stations in the deep level tubes or reconfigured them. I know that New York enlarged many platforms, but I don’t know if it was that complicated (i.e. lengthening the station ends by various means).
You’re also limited by operations – fresh in my mind, as I’ve been reading about the “cap” being built over the access tracks to Penn Station in New York and how they how one one hour window per week to be able to shut off power to the track level.
I’m thinking the whole shebang – drill rigs, concrete batch plant (yes, they make mobile ones as small as a plumber’s truck), crew shuttles, AND spoil removal. Plus, BART has a service window weeknights (~ 0100 – 0500) and weekends (~ 0100 – 0600 [Sat.]/ 0800 [Sun.]). The Colma Yard isn’t that far away (~ ten miles) so the trip would take about twenty minutes non-stop.
The main safety issue that I see is the third rail. The work crews are going to need electricity for their tools. So another service box will have to be spliced into the existing power lines without disabling the remote kill option for the nearest third rail segment.
What about ventilation and dust? You’d want to seal up the tunnels either end and then vent out – so you’d need a “man shaft” (and you’d need to do that anyways for elevators and stairs, etc down to the platforms). Interesting thoughts though.
This really is an intriguing idea.
BART has a lot of potential for infill in the East Bay, and judging by their Metro Vision plan for the next ~40 years, any sort of extensions will be in the SF-Oakland core area.
The hugely expensive outside platforms proposed for Montgomery and Embarcadero are a poor idea. BART’s next generation of cars will have 3 doors per side thus allowing much faster boarding/alighting. If BART can reduce headways to a reliable 2 min (currently 3-4 by schedule), and operate full length trains, then the station crowding should ease. Lengthening platforms system wide for 11 car trains would one suspects a better investment. Adding more stairs/escalators to the existing stations for greater access should also be considered.
I agree, the (next) new control system and additional doors should be reviewed before diving under the sidewalks/baseements of lower Market Street. But I sorta think BART would be tough to extend platforms, at least in San Francisco and Oakland. Most underground BART stations are book-ended by emergency stairs, ventilation fans, two cross-overs, one junction (or two including MacArthur) and at least one tight curve.
The cost of the station has been constantly repeated at $29 million, but that is incorrect. The total cost of the project was $57 million:
Maybe this is how they make transit projects in Europe look so cheap – only report the cost of actually building the final building!
St. Louis just won a TIGER grant that will permit them to add an infill station at Boyle Ave between Central West End and Grand stations.
“In Atlanta, similarly, the gap between Arts Center and Lindbergh Center on the Red and Gold Marta lines is 2.7 miles, passing through a zone of potential redevelopment.”
That’s another one that looks good on a map but has some issues. That area is a nifty small manufacturing district that’s attractive to innovative businesses, including the Sweetwater 420 Brewery. While there are other such areas in the city, if the area was redeveloped, it’s likely that at least some of the existing tenants would relocate to the ‘burbs. Moreover, the area has already been hard hit by MARTA’s decision to put a railyard there– and the railyard would complicate siting of an infill station. Beyond that, the area is isolated from the street network, with access only via the Armour Road interchange from old I-85 (which, BTW/IMO is the ugliest place in Atlanta!).
That said, if the rail transit element of the Beltline is ever built, it would be really nice to have a station there. We’ll see, I guess.
Believe it or not, the abortive recent transportation sales tax proposal included an infill station, not there, but on the Proctor Creek Green Line spur, which would’ve provided access to the Beltline and to the strange North Avenue Streetcar corridor that was a component of the sales tax program. The Beltline would also benefit greatly from another infill station on the South Line, plus some better arrangement on the East Line.
Dallas built an infill station recently as well.
Great article. Thanks for pointing out how Chicago has done a good job with adding in-line station. There are plans for some more in the next two years including one that will help conventioneers get to McCormick Place without needing a cab.
Seems like Metra in Chicago has plenty of inline station potential. Just as an example, the UP North line could be electrified, at least out to Evanston, with a third track in the middle for express commuter service. It would be a cheap metro-style line through a dense urban environment, while still retaining an express track for commuter service, with relatively low cost. Bring on the infill.
[…] Freemark at the Transport Politic calls them infill stations: new transit stops built in gaps along existing rail lines. Current […]
[…] conversion of railway lines to tramways can create opportunities to add new infill stations along the route (as trams have better acceleration so can stop/start more regularly whilst […]