South Station Expansion Project Discussion

But wouldn't longer turns actually work out better to pad for delays? Shorter turns = cascading delays if anything goes wrong. Why can't we have both the SSX AND try shorter turns? If it works, great, if it doesn't, then we'll still have more capacity at South Station.

The Governor of Maine's cousin must have had a hand in writing the South Station Comment..

diburning wrote:But wouldn't longer turns actually work out better to pad for delays? Shorter turns = cascading delays if anything goes wrong. Why can't we have both the SSX AND try shorter turns? If it works, great, if it doesn't, then we'll still have more capacity at South Station.

Transitmatters other point is that electric locomotives are 10x to 40x more reliable than the Ts diesels cutting the cause of delays and the need for padding--spend $2b on service that is 40% faster and 10x as reliable, not $2b on making schedule-padding easier.

diburning wrote:But wouldn't longer turns actually work out better to pad for delays? Shorter turns = cascading delays if anything goes wrong. Why can't we have both the SSX AND try shorter turns? If it works, great, if it doesn't, then we'll still have more capacity at South Station.

Seems to me the bar is low enough as it is. I wouldn't be in favor of lowering it any more. A little hustle might go a long way.

Padding is an expensive way of keeping trains on time: staffing sits idle, trains sit idle, platforms crowded with idle crews and trains. Imagine instead emus that ran faster, turned faster, and kept time (or made up lost time). That's worth $2b of infrastructure in a way that platforms for 35 min turns are not.

There are plenty of trains that turn at South Station in 15-20 minutes. Transit Matters says "average platform occupancy" which means it includes mid-day or late evening trains that don't need to turn fast for the next trip or clear the station for more trains. What they are trying to say is that there is room to add service with the existing capacity of South Station if you turn the trains around faster, but the way they presented the information slants it. The statement may be true, but not necessarily true during rush hour which is where you need the most amount of capacity. That is where the highest demand is and will always be - not matter what kind of ridership you generate with subway-style headways mid-day.

Even Dave's own example set has a 12-minute turn at South Station during the rush (with my emphasis)

Dave's Turn Table wrote:Depart Worcester as P500 at 4:45 AM
Arrive South Station 6:18 AM
Depart South Station as 6585 at 6:30 AM
Arrive Framingham at 7:00 AM
Depart Framingham as P586 at 7:15 AM

Layover yard space near South Station is basicly "Maxed out",Having yard space at Beacon Park for Worcester service or Widett Circle is needed.
Still need to add Platforms anyway If the "South Coast line" is finally built.
Time for the Post Office DC to move!

There are presently 13 tracks at BOS. When you get past the switches, there are about 9 main-line tracks (the NEC has 3 right? and 2 going towards the Old Colony branches?). Assume very efficient switching and very efficient swaps to the yards. Amtrak is going to occupy at least two tracks, one for Acela and one for Northeast Regional. Perhaps you allow two trains on some platforms to board at the same time. Still needs to be expanded.

Amtrak uses well over 2 tracks, and their service is more likely to increase rather than decrease, especially with the Acela 2. South Coast Rail and other proposals including service to Foxboro, Full time service to Buzzards Bay and part time service to the Cape, and the long term potential of Springfield-Boston service all require capacity in Boston. Diminishing layover times compromises the service. Crews need time to cover various administrative functions between trips (not to mention eat and use the bathroom), just because they’re not on a train doesn’t mean they’re “idle.” Trains also need to be cleaned and prepped for their next trip (while this doesn’t always happen, it should). What’s needed is flexibility, and a strangled terminal hub constricts that.

I believe NSRL is an overtired argument that our grandkids’ grandkids will be listening to 150+ years from now. Despite that, I believe BOTH North and South stations need expansion (to what they were before) with or without NSRL. The region needs both expanded service and increased service on existing lines. Thusly, it needs more capacity at its hub.

How about we clean trains and pad schedules at platforms at the trip ends where space is cheap?

Assign time-consuming tasks to the cheap hinterland platforms which do not have extreme operating pressures. If fast turns are possible only at one end, do the fast turn at South Station. The slow clean-and-pad turns should be moved to a platform in the hinterlands

Assume a system of 20 lines/termini. Some already have berths good for cleaning and schedule padding. Some will need new platforms. Some will need ADA vertical circulation (all of which Transitmatters has built into their $2b). In all, call it $100m in costs to certify, upgrade, or build 20 berths (one at each of 20 termini) at which 20 trains could simultaneously be berthed/cleaned/schedule-padded.

If lines are running at 15 or 30 min clockface frequencies, cleaners/inspectors/whomever at route ends would also keep plenty busy, all day.

Meanwhile at South Station, 4 new platforms and 8 new tracks and all vertical circulation to span such a big complex are budgeted at $2b/8 = $250m per track

Hinterlands = $5m per terminal berth
South Station = $250m per terminal berth

If time consuming terminal tasks are your justification for SSX, note that you're spending 50x to do them downtown. It is very hard to make that back on staff efficiency any task where you're already in that kind of capital hole.

Want to spend $500m on platforms/tracks to accommodate Amtrak's Acela 2 platform needs? Throw that into the NSRL pot and build them a facility in Woburn to turn trains from the NEC (which is where the expected growth is...not from the Lakeshore Limited or even 6/day Inland/Montreal).

Here's another way to look at how much dwelling in the core costs vs dwelling at line-ends:
Assume 250 workdays per year, a 40 year service life, and 10 peak-capacity hours* per day, implying that each new platform is being procured to handle the crush load during a 100,000 peak-hour service life.

8 platforms at South Station deliver 800,000 peak service hours for $2b = $2,500 per peak hour of platform occupancy
20 platforms at line termini deliver 1,000,000** peak service hours for $100m = $100 per peak hour of platform occupancy

Every time you can move a minute of platform dwell from the core to the fringe, you save $2400 in platform cost
Every time you can eliminate a minute of platform dwell from the core, you save $2500 in platform cost.

* Let's say that peak hours at South Station are 5am to 10am and 3pm to 8pm. The SSX is being justified by peak-hour needs, not by anything happening at 11a or 10p. If you shorten the hours, the dollars per hour will change, but not the ratio of 25:1 for the cost at SSX vs the cost of hinterland platform expansion.

** I'll even grant that platforms in the hinterlands only last half as long as one in the core due to service extensions (obsolescence/stranding) and exposure to the elements. There are good reasons to assume a platform's investment horizon is either 80 years at SSX vs 40 years in the hinterlands, or 40 years at SSX vs 20 years in the hinterlands

Maybe we should work on reducing the supposed $250M each platform cost at SSX to something more reasonable. The re-acquisition cost for the General Mail Facility should be high, but building the platforms and roofs should be single digit millions. No stairs, elevators, escalators. Throw in a cheapskate metal shed expansion of the waiting room; make is a quonset hut with LED lighting to class up the joint.

If land is so cheap at the remote termini, why hasn't the T done that already?

Arlington wrote:How about we clean trains and pad schedules at platforms at the trip ends where space is cheap?

Not a bad Idea for many of the commuter rail end points, though you will have to rework places like Haverhill or Providence. It looks like with new construction the end of the line is off the main like in Wachusett which would benefit your plan.

BandA wrote:If land is so cheap at the remote termini, why hasn't the T done that already?

Seems--based on Transitmatters' success at finding opportunities for improvement--that nobody at the T was asked to look at the problem, or not asked to look in a particular way.

Certainly was the case with the last E-train run of the night costing, what, $2m/year? (to handle an average of 1 rider), and seems like it was the case with EMU/Electrification.

The last E thing was complete nonsense. The last E train that used to connect (which was never actually the last train) is quite often at Park Street during "East and West", as it was when it was "costing" some nonsense amount that was conjured by some abstract number crunch factoring all kinds of things completely unrelated to the single trip. That fallacious report also continues to negatively impact numerous riders well in excess of the alleged 1 that get stranded in town when that train doesn't make it in time for the connection. You'd have to actually be there to see it, though, as data attainable online is completely void of such vital, practical, information. I say this not to get into that again, but to point out the lack of practical understanding that goes into this type of data analysis. There's a reason I don't make suggestions to UPS about how to handle their air parcel service, I have no idea how it works!

In breaking down the cost of the station expansion into dollars per hour, you factor a "40 year service life" for a station that's already 119 years old. You then use this arbitrary figure to make a comparison against something that is actual but not measurable in financial units. All this number crunching and hypothetical scenario analysis fails to take into account a ton of logistical realities that affect the way things actually work. Schedule padding at outlying points doesn't benefit anything save for a given train set's next trip, maybe. A set laying over in Worcester does nothing for a Providence train the equipment of which has still not made it in from Readville by departure time. The system needs flexibility at its hub, not where land value is the cheapest. You also have centralized functions consolidated into two division headquarters. Creating duplicate and redundant administrative facilities would quickly and vastly overshadow whatever efficiency benefits might be realized by rushing trains out of South Station.

Lastly, a train station needs a lot of space at the surface level. It has no use for nor conflict with the real estate above it. Such can be developed into commercial use generating perpetual revenue to the T and the Commonwealth alike, and the bigger the property, the greater the potential.

So there you have it, both practical and financial reasons to build the thing despite someone's opposition that from what I can tell equates to not buying a car so you can afford its gas.