Grand Junction Branch (The North/South Side Connection)

Discussion relating to commuter rail, light rail, and subway operations of the MBTA.

Moderators: CRail, sery2831

Re: Grand Junction Branch (The North/South Side Connection)

Postby BandA » Tue Oct 10, 2017 9:03 pm

Rubber tired road vehicles are superior to rail for stopping distance. They are worse for fuel efficiency, but that is nothing compared to labor & rolling stock cost. And rubber tires don't require PTC mandate or FRA regulations.

Can't use the Grand Junction for trolleys, it is already being used for freight and CR equipment moves.
User avatar
BandA
 
Posts: 1880
Joined: Thu Oct 11, 2012 11:47 am

Re: Grand Junction Branch (The North/South Side Connection)

Postby F-line to Dudley via Park » Tue Oct 10, 2017 9:12 pm

This isn't solely about cars. The 2012 CTPS study for WOR-BON peak-direction only commuter service also chalked up significant gates-down impacts to bicyclists and pedestrians at the Grand Junction crossings. That's a function of these crossings being right in the middle of MIT campus. Throw in the 1 + CT1 routes on Mass Ave. (>15,000 daily riders, two buses that carry over 50% as many riders as all of northside commuter rail put together), the CT2 + 64 + 68 + 85 (~7900 daily riders) on Broadway/Main, and all manner of quasi-private campus shuttles. Yes, the gate times at 15-minute bi-directional Indigo headways have pretty far-reaching impacts across many modes that share the crossings. And the riders--on all modes--negatively impacted by gate delays clocks in very high relative to the most optimistic ridership projections for an Indigo-ified Grand Junction shuttle.

The counts in the WOR-BON study aren't pretty...tolerable for the limited rush-only unidirectional nature of that proposed service, but fully indicative of the limitations of this line's functional throughput. Impacts were mainly self-mitigated by the limited hours of service and the unidirectional nature of service, which helped gerrymander those WOR-BON trains to least-impactful slots. For example, trains clear Mass Ave. faster going inbound-only in the AM off the Cambridgeport straightaway at 30+ MPH than PM when they're crawling outbound-only out of the Kendall station and sharp Main St. curve at barely >15 MPH, so it proved fair all-around to have that crossing's severest impacts limited to a narrow 3-hour afternoon window because mornings wouldn't be too bad. The line being totally idle for the entirety of the midday off-peak also spared a lot of ped/bike impacts, which have hourly surges when MIT students change classes and a big lunchtime surge at all the Kendall offices. It would be totally different if you went all-day, bi-directional, with the "gold-standard urban rail" 15-20 minute frequency targets each direction that more or less define where Purple Line becomes real-deal Indigo Line. It looks great on a 2D map, but very ugly on the multi-multi-modal math. And Gov. Patrick's magic 2024 Indigo spider map that showed this line on the network didn't even attempt to update the CTPS math; it was included because it looked great on a 2D map.


It's extremely difficult to justify a build that has such immediate all-day negatives to all other modes across the spectrum. Modal warfare never ends up making for good transit decisions. That's true when bustitutions have been rammed down the throat of well-functioning rail routes, and also unfortunately true of making every varied user of the road bow at extremely congested crossings to levels of RR traffic that have never been seen in 120 years on this particular line. The decision can't come down to picking winners and losers amongst existing transit routes. #1 bus riders, more numerous than the potential riders of this shuttle and with NO rail augmentations for where they need to go, can't just be asked to suck it up because they're wearing the same less aesthetically pleasing sets of wheels as the dastardly cars on that road.


All of that is endemic to the mainline RR mode having the right of way all the time at crossings. The most precision-tech traffic signal prioritization in the world and most precision-timed departures in the world from the flanking CR stations are only going to do so much to manage gate queues when the trains can't be preempted by any other mode by federal law of the land. At most optimistic the prioritization dance just keeps the jumble of modes from falling off a very thin tightrope; they'll always be teetering unsteadily on that tightrope. To do this truly right at frequencies worth a damn you need a mode that can outright share signal phases at the crossings. And that's not something that shares FRA rails. That's BRT (boo) or LRT (yay!)...the very modes studied on the OFFICIAL Urban Ring proposal studied for this route. It's got huge up-front cost (though perhaps less than the official study if the $B's cross-Brookline tunnel is traded in for a BU Bridge subway extension under the reservation), but there's no question it delivers the advertised frequencies that draw the advertised ridership while being able to share fully-distributed signal phases at the crossings. There isn't a DMU/EMU Indigo kludge that'll do that without making many other crossing modes pay a heavy price.
Last edited by F-line to Dudley via Park on Tue Oct 10, 2017 10:00 pm, edited 1 time in total.
F-line to Dudley via Park
 
Posts: 7184
Joined: Fri Mar 11, 2005 7:26 pm
Location: North Cambridge

Re: Grand Junction Branch (The North/South Side Connection)

Postby F-line to Dudley via Park » Tue Oct 10, 2017 9:54 pm

BandA wrote:Can't use the Grand Junction for trolleys, it is already being used for freight and CR equipment moves.


Not true. You can take it off the RR network without unduly impacting commuter rail use if certain prerequisites are met:

1. More equipment independence between northside and southside. If the T had enough padding on its reserves, there wouldn't be a need to keep shuttling cars to keep equipment availability on each side in such precise balance in the face of yearly shortages. Same goes for work equipment, which they are starved for systemwide and have to triage overwhelmingly from northside.

2. More places to store that equipment. If they luck out on that Widett Circle real estate deal with the bottom-level storage easement they're pretty much set here...and set for freeing up some reserve work equipment space at their (incredible shrinking) Beacon Park easement.

3. Full-service maintenance facility for the southside. A start on electrifying the Providence + Indigo-Fairmount + Indigo-Riverside lines kicks off that process by default, since the EMU's or electric locos would be captive south-only. You'd also want that same facility to be able to handle day-to-day coach servicing needs to start differentiating S&I jobs with BET. Say...diesel locos still have all their work done north and any heavy-repair jobs go north, but southside gets fortified to keep up on inspections and day-to-day maintenance of the coach fleet.

4. The PAR Worcester Branch is upgraded to Class 2 or 3 speeds at full state-of-repair. This is going to happen anyway with a more competent owner than PAR, because once NS shears off the Patriot Corridor then Worcester-Portland becomes 'the' freight main. MassDOT's going to burn some cash helping out with that rebuild in the next decade for the sake of the next freight regime, so eventually the CR Hospital Trains will be able to sustain 35+ MPH and get the round-trip done in under 3 instead of 5+ hours.

5. All of the above in #1-3--fortifying southside equipment reserves + storage, adding southside maint facilities + differentiation with BET--limits the required number of north-south equipment swaps from up to 2 a day to only ~3 per week. That keeps the operating costs of running much longer-distance in *relative* balance with today. And as per #3, the swaps are going to lean heavier on shuffling diesel power and lighter on shuffling coaches because southside is picking up more S&I tasks and starting to differentiate its fleet with initial introduction of electrics. So it may cost more per trip, and CR employees might grumble that it's a lot less convenient...but *relative* equilibrium is maintained systemwide so the re-route doesn't become any undue burden.


Freight will probably solve itself. CSX stops serving Houghton Chemical @ Beacon Park on its Everett daily come July 2018, which opens up an opportunity to shove their Everett customers off on a haulage deal to PAR. CSX can keep their customers and chase new business at Everett Terminal, but PAR pockets some extra cash by tacking CSX's loads onto their own Everett BO-# freight and interchanging with CSX at Worcester...while CSX makes more total money keeping the loads but saving more in cut staff and operating costs by abolishing the Everett job. So long as PAR meets on-time performance targets at the interchange, the loads arrive in Framingham all the same...only now they're picked up on the westbound daily that handles the G&U + P&W + PAR interchanges instead of on the eastbound direct to Everett. In the event CSX ever needs to take back the job for itself, the state can substitute an overhead rights agreement on the Fitchburg Line inbound of Ayer in place of the deleted Grand Junction (GJ has a system-worst Plate B freight clearance because of the very low Memorial Drive overpass, so Plate C on the inner Fitchburg actually ends up a capacity improvement for CSX).



^^This (excepting the carrier-to-carrier freight outsourcing which will probably happen within 18 months) is several hundred $M's in eat-your-peas stuff. But it's all upgrades the state has to do anyway to keep up with general growth: the equipment reserves, the southside storage, the southside maint facilities, the upgraded freight line. So all you have to do is group these loosely into a master plan instead of treating them each as wholly independent bucket-list items, and start plugging away. Past a certain fortified threshold, you're A-OK to take the Grand Junction even without an intra-city replacement like North-South Rail Link advancing any further beyond the next study-about-studies. The upside in being able to string together real rapid transit on that route so insanely outweighs the most convenient Hospital Train route that it's a no-brainer. Just don't leap until you've eaten enough peas to not need to run the Hospital Train more than a few times a week for sanity's sake.

As described earlier from the CTPS study, the only reason demand for a Worcester-North Station direct exists for even part of the day is because Red and Orange are too oft-FUBAR'ed at rush hour. Fix the subway state-of-repair and fix the dwell problems that are choking the Big Four downtown transfer stations, and the patronage for the B&A one-seat to Kendall & BON vanishes. The only route of any kind that gets outright prevented by taking the GJ is an Amtrak New York-Portland daily via the Inland Route and a North Station reverse. And that's not more than a tiny niche of thru riders once you take away the Boston overchurn of typical Inland Route patronage vs. typical Downeaster patronage. Nothing you'd ever want to blink on a rapid transit build by fretting over, especially if fixing Orange makes the current BBY-BON transfer leg between NEC trains and the Downeaster a lot more real-world useful.
F-line to Dudley via Park
 
Posts: 7184
Joined: Fri Mar 11, 2005 7:26 pm
Location: North Cambridge

Re: Grand Junction Branch (The North/South Side Connection)

Postby BandA » Tue Oct 10, 2017 11:15 pm

AFAIK there are no plans to build a real south-side maintenance facility. Is there a project for that? How many $B? And the stated reason for WOR-BON is that BOS is full. Until the south station expansion is built, which requires real estate purchase that hasn't happened. [sarcasm]If BOS is full, they could always add more trains to the Worcester line and truncate them at BBY, and everyone can hop onto the Orange Line[/sarcasm]

Or they can fix the grade crossings in Cambridge. Grade separate some of them, make the gate closings shorter / more intelligent. Restrict the speed of the CR trains on the branch, which wouldn't be too bad. Since the train is going slow, might as well stop at both Kendall and Cambridge St. I'm leaving the solution to routing the eastern end as a future exercise.
User avatar
BandA
 
Posts: 1880
Joined: Thu Oct 11, 2012 11:47 am

Re: Grand Junction Branch (The North/South Side Connection)

Postby F-line to Dudley via Park » Wed Oct 11, 2017 7:25 am

As I said...if any SS lines go electric, you don't have a choice but to maintain them down there with a new facility because the equipment becomes captive to the south and loaning out to Amtrak Southampton shop is out of the question when they're at-capacity. So that's a self-correcting problem by doing the right thing and running electric on Providence + the Indigos. They aren't at a loss for space at Readville for shops in Yards 2 or 5 if they can close any of the proposed land deals under negotiation for more storage near the terminal. All of those bucket list items are self-correcting by doing what they have to do for strictly general-purpose growth.

-------------------------

Eliminating the worst GJ crossings is physically impossible on RR mode because of maximum recommended grades. Main St. has the air rights overhang and the Red Line traveling below...can't go over or under with the road or the rails. Broadway is too close to Main at only 800 ft. away for maximum grades to go up/down with the rails, and the road is too closely abutted by towers to change the elevation of the roadway. Mass Ave. is too close at 950 ft. to the sharp curve running into Main to drop the rails in-time, and there's the power plant air rights overhang only 400 ft. away preventing a rise.

There was many years ago an old MIT Master Plan concept for sinking Mass Ave. thru traffic under a pedestrian plaza from Amherst to Windsor St.'s sort of like Harvard's cover-over of Cambridge St. near Mass Ave. That would've buried the road under the rails and Albany/Vassar intersections for thru *car* traffic, but kept a one-way pair of single-lane pop-up frontages for Albany/Vassar and the buses...so only a partial elimination for the one personal-transit mode we're least enthusiastic about feeding $$$ with more urban megaprojects. The problem was it would've cost a kajillion dollars on both public and private ends for what amounted to a campus vanity project, and was a dubious waterproofing prospect that close to Charles Basin since all of Mass Ave. south of Albany was old bay water landfilled from 1905-10 (the Grand Junction pre-fill used to run through Cambridgeport on a riverfront causeway from Waverley St. to Main). And the buses/bikes/pedestrians still would've gotten stuck gates-down half the day on the surface all the same. It's almost certainly a nonstarter on its engineering merits, and the amount of money MassDOT would have to contribute is probably enough to get two-thirds of the Comm Ave. subway extension dug out under the B reservation from Blandford to BU Bridge for making the GJ mode conversion to trolley in the first place...so basically robbing from the coffers of more effective builds to give MIT a pretty plaza and thru cars an insta-skip of 2 lights.

Cambridge St. and Medford St. are simply least concerns compared to the Big Three of Mass Ave./Main/Broadway...so if you have no non-kludgy solutions for the crossings that are causing all the trouble it's not going to gain you anything to spend excess cash eliminating the others for perfectionism's sake. The only semi-expendable crossing is Binney, which has low volumes and can probably be outright closed as a thru street between Galileo Galilei Way and Cardinal Medeiros in favor of converting into a pedestrian crossing / part-time service driveway. In fact, they'd probably do exactly that conversion for Binney if the limited WOR-BON rush-hour proposal got put back on the front-burner.


Again...you need a mode that can outright share signal phases because some crossings like Main aren't going away in any scenario, and Broadway 800 ft. away is tough to picture going in any scenario if Main can't. And you need a mode that can climb steeper grades if you want to get rid of Mass Ave. BRT and LRT can meet both conditions at all 3 problem crossings, with LRT being hands-down more advantageous than BRT at climbing the steepest grades within the narrowest ROW footprint at the lowest construction costs for creating those grades. It can't physically be done on an FRA RR, and the partial mitigations like MIT's vanity Mass Ave. burial just shovel insane sums of money at cars-first instead of actually focusing all attention on the multi-modal transit problem. So use the right tool from the toolbox. Trying to force-fit an Indigo route here hell or high water is the very definition of modal warfare. It dumps on all other modes so some planner's rooting interest in mode can be "proved <bleeping> right" on its own proof-of-concept, which is the antithesis of actually solving a transit problem. It's making the same mistake of planning stubbornness that sank the Silver Line Phase III BRT tunnel under Boston Common on its own largesse and sacked Roxbury with an inferior bus instead of the "equal-or-better" replacement rail they were promised.

When that ends up the sole aim, it usually doesn't follow through as real-world cheaper to spend less on broken transit. We found out with Silver-Washington that what you don't end up spending on a signature project making useful baseline transit just ends up getting wasted on frivolous amenities and aesthetics to paper over the fact that they used the wrong tools for the job...i.e. no boondoggle too big to "prove <bleeping> right" somebody's choice of modal warfare. Since the Big Three crossings make any Indigo-frequency urban rail route broken by design for its unfavorable multimodal impacts, there is pretty much no practical political way of mounting the effort that wouldn't end up overdosing on frivolous ancillary costs as a means of self-justifying the mode choice. A cheap DMU to cheap bare-bones platforms that ends up tanking the OTP of the already barebones bus routes which carry far more riders isn't going to be acceptable to anyone...but magnificent station overbuilds? Hell yeah, MIT already showed with that kooky Mass Ave. plaza deck-over proposal that they're up for grandiose monument-building on the public dime. So that ends up being the terms of engagement for "mitigating" the modal choice's downsides. Unfortunately in public works that counts as prevailing logic, so the plot was lost the second the wrong tool was preselected for the job.
F-line to Dudley via Park
 
Posts: 7184
Joined: Fri Mar 11, 2005 7:26 pm
Location: North Cambridge

Re: Grand Junction Branch (The North/South Side Connection)

Postby BandA » Wed Oct 11, 2017 4:07 pm

How long do the crossing gates need to be down before a train that is travelling 10MPH? How long would it take an 8-car train to clear a crossing if it is starting from a stop that is adjacent to the crossing?
User avatar
BandA
 
Posts: 1880
Joined: Thu Oct 11, 2012 11:47 am

Re: Grand Junction Branch (The North/South Side Connection)

Postby F-line to Dudley via Park » Wed Oct 11, 2017 5:35 pm

BandA wrote:How long do the crossing gates need to be down before a train that is travelling 10MPH? How long would it take an 8-car train to clear a crossing if it is starting from a stop that is adjacent to the crossing?


The CTPS study cites that, so you can look on the PDF linky a few posts up. There's one calculation on p. 62 that breaks down the FRA-regulated standard gate times from the train's perspective through the whole protection sequence: track circuit activation --> flashers activate --> gates drop --> train occupies --> track circuit deactivation --> gates lift --> flashers deactivate. The Kendall station stop, since it would have a platform adjacent to the Main and/or Broadway grade crossings, would have a DTMF switch so gates stay up and road traffic moves through the duration of the station stop. Then the engineer presses a button on the control stand sending a signal activating the gates, and only then will the train proceed. It's a +/- 20 second difference in total protection time between going through a crossing in-motion at 25-30 MPH vs. triggering the DTMF switch from a dead stop then accelerating through. For the Big Three crossings the protection behind the platform would run 63 seconds because the train is decelerating in-motion, while the crossings ahead of the platform would be 81-83 seconds because of the DTMF lag + acceleration out of a dead stop. If, say, the platform were placed under the air rights overhang Mass Ave. crossing protection would run 63 secs. in the inbound direction but 83 outbound...while Main + Broadway would each run 83 inbound and 63 outbound. Adjust accordingly if you place the station at the mid-block between Main/Broadway instead of under the overhang...then the grouping changes a little bit: Main/Mass = 61-63 inbound/81-83 outbound, Broadway = 81-83 inbound/63 outbound.

Note that because these WOR-BON trips are unidirectional by peak Mass Ave. only takes it really heavy on the PM commute, and comes out relatively unscathed on the AM with a 20-second quicker protection cycle. That was one of the key reasons the study checked out on tolerable impact: the "worsts" at each crossing were contained to one 3-hour period daily, and there was a time-of-day split divvying up when each of the Big Three had their "worsts"...meaning half of Cambridge wouldn't get locked up because at least 1 of the 3 crossings would be on the 20 secs. quicker direction of train traffic throughout any given peak. Bi-directional/reverse-commute service would've been much thornier if it made all crossings equally bad on both peaks.



Ch. 5 details the on-street vehicle impacts, with charts calculating best-case to worst-case traffic delays by vehicles-per-hour delayed at each crossing, and best-case to worst-case delay time to road traffic at each crossing. Note that this is a different calculation than the train-only protection cycle calculations detailed above because of the extra lag required for accumulating then dumping stopped car queues for each crossing event. And note that just as before there's a 20-second inbound train vs. outbound train difference in length of protection cycles due to train speed coming in or out of the Kendall stop, so road traffic likewise has a +/- 20 sec. differential by train direction...then padded by several seconds for the lag in car queue reaction time.

Mass Ave. traffic was stopped anywhere from 2:30 to just under 3:00 (IB vs. OB) minutes under the best-case traffic modeling, and 3:30 / >4:00 under the worst-case modeling . Broadway was worse at 2:30 / ~3:10 under best-case, 4:00 / ~4:15 under worst-case traffic modeling. Main comparably better at ~2:20 / 2:40 best-case and ~3:10 / ~3:40 worst-case, but a bigger directional split because the presumed platform location was immediately adjacent and had to contend for the DTMF activation lag in the inbound direction.

You can then extrapolate from here using the veh/hr. delay charts and time-of-day data from Ch. 3 to calculate the size of the traffic jams. Similar counts available for bike and ped traffic by time-of-day, and the buses are listed by ridership and headways by time-of-day so rougher math is calculable for delays on transit and ped/bike shares.


Pretty ugly all-around, but within tolerability for limited unidirectional peak-only service because each crossing trades off differing easier peaks vs. worse peak by IB/OB direction to constrain the ill effects. But you can easily see how going full-on bidirectional makes things a lot harder, cranking up the frequencies to Indigo levels REALLY REALLY gums up the works, and adding additional infill station stops at Mass Ave. for denser rapid transit-like stop spacing becomes completely untenable by slowing trains down and baking in more DTMF signal trigger lags. The FRA's also unyielding on how many seconds the protection cycle must run its course upon activation/deactivation, so it's not like you can use faster-raising gate motors or cut corners on the lag between track circuit trigger flashers on/off to shave precious seconds, like gated LRT/BRT crossings in other parts of the country are able to. There's a reason why all RR gates statewide drop and raise at more or less the same sluggish speed while the gates protecting the Silver Line D St. portal are so much quicker up/down; it's a regulatory, not technological, difference.
F-line to Dudley via Park
 
Posts: 7184
Joined: Fri Mar 11, 2005 7:26 pm
Location: North Cambridge

Re: Grand Junction Branch (The North/South Side Connection)

Postby CRail » Thu Oct 12, 2017 1:50 pm

The number of hours in a given day do not avail themselves to reading the novels above. So, admittedly, I've not. I did read the first one subsequent to my previous post. Regarding the impacts of 15 minute bi-directional headways all day long and crossing priority, yadda yadda.

First of all, 15 minute headways all day? What are you nuts!? We aren't talking about running rapid transit on the surface, we're talking about a commuter option in which some service takes an alternate route. I envision a few trips through there during the rush when no one is getting through there in a timely manner anyhow and maybe a couple mid day trips. This isn't providing a service that's not already there, it's adding a direct connection in an effort to get more cars off the street in a continuously congested area. Did the factors in the impact study include trains not crawling through at 10mph?

Regarding the "all hail the train" grade crossing attitude nonsense, it's completely nonfactual. Plenty of railroad crossings exist even on our very own system whereat the train gets a Stop Signal until the gates have finished closing. Some crossings require a train to activate it when it's ready to cross, and the train is held until the crossing is ready. So long as the train has advanced warning that it will have to stop at the crossing (a distant signal at Approach), the train does not need to have absolute priority.
Moderator: Massachusetts Bay Transportation Authority
Avatar:3679A (since wrecked)/3623B (now in service as 3636B).
User avatar
CRail
 
Posts: 2131
Joined: Tue May 18, 2004 8:27 am
Location: Eastie

Previous

Return to Massachusetts Bay Transportation Authority (MBTA)

Who is online

Users browsing this forum: Google Feedfetcher and 4 guests