Green Line Extension to Medford & Union Sq

The effect on the station designs moving from prepayment to proof of payment would be interesting, but I suspect we are talking about millions in cost savings and not tens of millions.

Remember, this is still a light rail system that people are treating like heavy rail for the GLX design. Take a look at some of the newer and very successful LRT systems like Dallas, Houston, Salt Lake, and Minnesota. The stations are a simple shelter and a slightly raised slab of concrete. The GLX stations in the design are larger and more involved than many HRT stations. The stations need to equate to the mode. PoP is now best practice for LRT.
 
Remember, this is still a light rail system that people are treating like heavy rail for the GLX design. Take a look at some of the newer and very successful LRT systems like Dallas, Houston, Salt Lake, and Minnesota. The stations are a simple shelter and a slightly raised slab of concrete. The GLX stations in the design are larger and more involved than many HRT stations. The stations need to equate to the mode. PoP is now best practice for LRT.
Yes, this. Google (Dallas) Dart Station Design or Houston LIght rail station design.

Ball Sq could have easily been a flat, cross-the-tracks-yourself kind of station with no headhouse. Lowell could tie straight into the bike paths. In fact, the Community Path could have done all the grade-changes and dumped everyone trackside inbound, and, yes, in the evenings, you'd have to cross the tracks on foot (and in ADA compliant crossings). No elevators needed.

College Ave & Brickbottom & Lechmere needs the full mechanicals, but I'd say Union Sq would work fine with level crossings and ramps.
 
^

Faster ops are the key being able to use the existing fleet (no Type 9.5 or Type 10 procurement) a.

The $118 million contract for 24 Type 9s to supply the additional equipment for GLX was already placed with CAF back in May 2014. If the MBTA were to cancel the order now after CAF has already begun final design and ordering components from suppliers for assembly, a large portion of that amount, if not all, would still have to be paid to CAF.
 
The $118 million contract for 24 Type 9s to supply the additional equipment for GLX was already placed with CAF back in May 2014.
That's fine. They can be used to run 3-car ops everywhere (and maybe some 4s). And/or retire the worst of the 7s or 8s. They will not go to waste, strictly speaking.

They are waste when you only buy them because you had to because you intend to have them move slow.

Or when you have to build a $300m yard to park them overnight, which is the case with the GLX.

If we can move the vehicles we have faster and cover an expanded system with a fleet that fits in existing yards, maybe just put the worst spillover vehicles (unrebuilt 7s?, early-retire some 8s?) in off-site storage as spares.

(Am I wrong in thinking these only covered Brickbottom/Union/Lechmere and increased frequencies? I thought there was another order/options for Type 9.5s that was going to be needed to fully populate the GLX)
 
Something else I'm curious about: would moving to a PoP system make it markedly easier to incorporate either a) across the board fare increases? or b) implement a varied pricing regime? (commuted fares for certain income levels, students, etc... I had an awesome discount on the Danish train system - 50% for regionals and inter-cities - because I was under-26 and a student and there was a specific mobile app where I purchase these reduced tickets from. I'd show up at the station, pull out my phone, by a ticket and a seat reservation, and do it all for half price, it was annoyingly simple).

It certainly provides the MBTA with flexibility to experiment, but it's worth considering in this discussion that Green Line is light-rail system in the US with the highest fare recovery ratio: close to 60% based on 2012 NTD data. The next two are in Portland and SD, which, when combined, are still lower in ridership than the GL. I know fare increases are in now way a panacea, but I think it was Arlington who brought a while ago in another thread that the 2007 fare increases had little impact on ridership (it grew post-increase even), which could indicate that the rapid transit system is an under-valued asset.
 
That's fine. They can be used to run 3-car ops everywhere (and maybe some 4s). And/or retire the worst of the 7s or 8s. They will not go to waste, strictly speaking.

T)

But that doesn't free up any money to go toward paying for GLX, replacement of the Type 7s and Type 8s is something that isn't funded yet, and only would have happened after GLX was built.

These 24 cars are for the full GLX, Union Sq only interim service was supposed to be covered by repairing out of service Type 8s.
 
But that doesn't free up any money to go toward paying for GLX, replacement of the Type 7s and Type 8s is something that isn't funded yet, and only would have happened after GLX was built.

These 24 cars are for the full GLX, Union Sq only interim service was supposed to be covered by repairing out of service Type 8s.

Ok, so we can't save $118m right now. But note that we *could have* with POP, we will (at the next procurement cycle) with POP, and all Billion worth of rolling stock that we *do* have will work better with POP.

If a Billlion dollars worth of rolling stock (186 x $5m = $930m) works 10% better/faster, it produces a non-cash savings or benefit worth about $100m per year *every year* Or fully-loaded with maintenance and storage costs, it probably only has to work 5% better to work $100m/year "better"

That the $118m can't be saved (neither can the $40m cost of flyover steel) is irrelevant to whether there's still about $300m too much worth of yard costs and $500m too much worth of station costs in the GLX because of bad thinking about how stations and vehicle floorplate and vehicle dwell time (elsewhere) are used to collect fares.

PS: if there's no need to watch a fare box on any cars in a train, that leaves the drivers "in back" just doing doors, which probably means we can use cameras/sensors/monitors to go to OPTO on 2, 3, and 4 car trains, both on-street and in-station like other systems have.
 
Looking at the numbers it seems it's not even worth arguing about shaving $100m for simpler stations. The cost is for the entire project, for materials and construction. This isn't an issue of added bloat just straight up miscalculation of costs. The stations as designed aren't as lavish as you guys seem to make them out to be. Sure they could be simpler but clearly that won't have enough of an effect of the project to worry about it.

If the state can't find the extra billion they aren't going to worry about simpler stations, they are just going to build a much smaller extension probably just to Union Sq and Washington St with a smaller train yard that could be expanded later on.
 
The difference between LRT-standard platform-with rain shelter vs HRT-standard stations is easily $40m/stop (in pre-overrun costs, pushing $80 in overrun mode) basic platforms x 6 stations =~ $300m to $500m (and is where the first big jump in costs from $800m to $1.3b came from back in 2007 or so.

In the earliest versions, it was "D Branch" and $800M ish. All the bloat since then has been in headhouses and yards. (And a bit extra for the Union Sq branch)
 
Looking at the numbers it seems it's not even worth arguing about shaving $100m for simpler stations. The cost is for the entire project, for materials and construction. This isn't an issue of added bloat just straight up miscalculation of costs. The stations as designed aren't as lavish as you guys seem to make them out to be. Sure they could be simpler but clearly that won't have enough of an effect of the project to worry about it.

If the state can't find the extra billion they aren't going to worry about simpler stations, they are just going to build a much smaller extension probably just to Union Sq and Washington St with a smaller train yard that could be expanded later on.

Van -- I think you have probably outlined what will come from the re-evaluation

The critical need is to get to Union Sq. where a lot of development can happen in the next couple of decades

Tufts already has just off campus access to Davis -- they can always run from College & Boston to /from Davis -- just like a Rt-128 business council bus [several sites to / from Alewife] with high frequency during the critical hours
 
The difference between LRT-standard platform-with rain shelter vs HRT-standard stations is easily $40m/stop (in pre-overrun costs, pushing $80 in overrun mode) basic platforms x 6 stations =~ $300m to $500m (and is where the first big jump in costs from $800m to $1.3b came from back in 2007 or so.

In the earliest versions, it was "D Branch" and $800M ish. All the bloat since then has been in headhouses and yards. (And a bit extra for the Union Sq branch)

Arlington -- outside of some extras such as noise screening -- I think that you have the analysis down of what happened to the GLX

its called -- Creeping Featurism a corollary of infinite project scope -- the result of expecting to not have to care how much it costs to build since its somebody else's money
 
This, from Norfolk's new system, is what an elevated station looks like under POP
IMG_0485.JPG
,
and this is what a surface station looks like
IMG_0427.JPG

contrast that with the BIG piles of concrete STRUCTURES that the GLX is getting. Want a 4-car station? Make it longer (not heavier, taller, or more mechanicals)

Arlington -- have you a $ figure for comparison of Norfolk's Light Rail to the GLX?

It would be nice to compare:

  • cost per mile
  • cost per station
  • cost per person

I'm willing to bet that the GLX is at least 150% of the Norfolk System
 
Cost per person might not be as bad as the cost per mile and cost per station comparison but the other stuff I think would have to be much higher than Norfolks. That is kind of a well duh statement though.
 
Taking a quick look at the current floor plans for most of the stations, it appears most stations have the same basic layout. There is 1) a site specific entry, 2) a large lobby with Charlie card vending machines and gates, 3) a very long corridor with 2 elevators, escalators, and stairs, and then 4) the platform below the street level. Everything is arranged in a linear fashion starting with the street entry and ending with the platform. Everything is between the inbound and outbound tracks. It looks like for these stations part 3 could be replaced with a 1 to 12 slope ramp. This would save a lot of money and would not require crossing tracks or POP. The stations could be further simplied by eliminating much of the walls, roofs and the restrooms. The roof could be a simple cover over the entry and ramp to help save snow removal costs.
 
Everything is between the inbound and outbound tracks. It looks like for these stations part 3 could be replaced with a 1 to 12 slope ramp. This would save a lot of money and would not require crossing tracks or POP.

As I stated previously, the ramps would be enormous. To comply with ADA, in addition to having a max slope of 1:12, you need a landing of 5' for every 30' of run.

The rise overall is probably going to be 15' or so (at least), meaning you would need at least 180' of run. 180'/30'=6 landings x 5' = +30' of landings. 180' + 30' = 210' total length of the ramp. A U-shaped ramp would have each segment be approx 105' long and the total footprint of the ramp would be at least 8' wide.
 
Complete phase 2/2a, run track to College Ave. Use Tufts to help fund that station. Leave Gilman, Lowel and ball for the moment. Put together new designs and financing plan for those stations. Financing could include increase in Somerville property tax as well as hike in T charges. Ideally it would be gas tax but that won't happen. When there's a proper plan in place, revisit the remaining 3 stations.

This seems eminently reasonable. Tufts clearly wants a green line stop and is already investing in new buildings and TOD. Furthermore, a Tufts (or further) stop is key to shortening and speeding up bus routes in Medford and farther Somerville. The bus realignment in and of itself would give a nice ridership boost to the MBTA rather than just transferring existing riders from bus to T.

Looking at the map, is one of the mid Somerville stations ideally placed to help rationalize bus routes within Somerville?

Overall, US transit has far too many stops. If they are the drivers of cost, space them further apart and cut the rest until the community can pay for them.
 
This seems eminently reasonable. Tufts clearly wants a green line stop and is already investing in new buildings and TOD. Furthermore, a Tufts (or further) stop is key to shortening and speeding up bus routes in Medford and farther Somerville. The bus realignment in and of itself would give a nice ridership boost to the MBTA rather than just transferring existing riders from bus to T.

Looking at the map, is one of the mid Somerville stations ideally placed to help rationalize bus routes within Somerville?

Overall, US transit has far too many stops. If they are the drivers of cost, space them further apart and cut the rest until the community can pay for them.

So basically just make either Somerville or a new TOD developer pay for half or more of the stations. And in the meantime just roll through. Quite a shakedown, but I like it.
 
All four of the following stations can be replaced with street-level access on the inbound platform and fully-ADA compliant access in which users cross the inbound tracks to get to/from the outbound platform

Ball Square (eliminate headhouse and Broadway Access)
Lowell Street (use community path for access)
Gilman Square (use community path extension for access)
Union Square (already has a platform-level plaza)

Saves so much money, you almost shouldn't care whether fares are collected or not. But if you pick POP, you can hire years and years of random checks with the money you save.

Retain headhouses at College Ave, and Washington St due to odd elevation changes. Lechmere to remain elevated.
 
What flew for acceptable rapid transit in 1959 does not fly in 2019. What works in Norfolk with a few hundred passengers a day does not work in Somerville with 5,000 passengers per day at stations. This is not a surface streetcar line, and this is not Newton with nicely sized residential lots and light walk-up traffic.

This is Somerville, with triple-decker density surrounding almost all the stations. Every single station is going to get passenger loads 2 to 4 times the typical D Branch station, much closer to actual heavy rail passenger counts. So you need to build your stations to accommodate those loads. That means grade separation, no bullshit with track crossings. They are not safe for passenger loads that high, they are less safe for people in wheelchairs, and having to wait for floods of people to cross the track will absolutely murder your schedule reliability.

The D Branch is not a good comparison whatsoever. Only five stations actually force track crossings (Brookline Hills, Beaconsfield, Newton Highlands, Waban, Woodland - all less than 2000 riders/day), and all to the inbound platform. The others have an entrance/exit to both platforms. That means that when you have the largest crowds at once, many of them can directly exit the platform. That simply would not be true at Ball Square and Lowell Street, plus Union Square (island platform needed as a terminus without room for tail tracks).

The vehicle is the technology, not the service. The service that Somerville demands and requires is rapid transit, regardless of whether that comes as Green Line, Orange Line, or Silver Line. That means a service where passengers are separated from the paths of the trains at all times, with station designs suited to handle massive crowds. The south end of the Orange Line is a much better comparison for what level of demand we're looking at. Those concrete bunkers wouldn't be any cheaper than the stations as they are currently designed, especially if the MBTA got sued again for accessibility requirements. Ramps are not suitable for actual mass access, and dual elevators are the name of the game. POP saves absolutely nothing for these stations except the actual faregates - you can't actually skimp out on much else without making the stations less usable, less safe, or less likely to pass code. By the time you go through literally years of redesign - in the process completely fucking over the community through delays and through stations that don't serve them as well as these do - you're not going to save any money no matter how much you fantasize about third-world asphalt platforms.
 
What flew for acceptable rapid transit in 1959 does not fly in 2019. What works in Norfolk with a few hundred passengers a day does not work in Somerville with 5,000 passengers per day at stations. This is not a surface streetcar line, and this is not Newton with nicely sized residential lots and light walk-up traffic.

This is Somerville, with triple-decker density surrounding almost all the stations. Every single station is going to get passenger loads 2 to 4 times the typical D Branch station, much closer to actual heavy rail passenger counts. So you need to build your stations to accommodate those loads. That means grade separation, no bullshit with track crossings. They are not safe for passenger loads that high, they are less safe for people in wheelchairs, and having to wait for floods of people to cross the track will absolutely murder your schedule reliability.

The D Branch is not a good comparison whatsoever. Only five stations actually force track crossings (Brookline Hills, Beaconsfield, Newton Highlands, Waban, Woodland - all less than 2000 riders/day), and all to the inbound platform. The others have an entrance/exit to both platforms. That means that when you have the largest crowds at once, many of them can directly exit the platform. That simply would not be true at Ball Square and Lowell Street, plus Union Square (island platform needed as a terminus without room for tail tracks).

The vehicle is the technology, not the service. The service that Somerville demands and requires is rapid transit, regardless of whether that comes as Green Line, Orange Line, or Silver Line. That means a service where passengers are separated from the paths of the trains at all times, with station designs suited to handle massive crowds. The south end of the Orange Line is a much better comparison for what level of demand we're looking at. Those concrete bunkers wouldn't be any cheaper than the stations as they are currently designed, especially if the MBTA got sued again for accessibility requirements. Ramps are not suitable for actual mass access, and dual elevators are the name of the game. POP saves absolutely nothing for these stations except the actual faregates - you can't actually skimp out on much else without making the stations less usable, less safe, or less likely to pass code. By the time you go through literally years of redesign - in the process completely fucking over the community through delays and through stations that don't serve them as well as these do - you're not going to save any money no matter how much you fantasize about third-world asphalt platforms.

I have nothing to add, but this - Well Said!
 

Back
Top