RR503

On 1/15/2021 at 3:15 PM, VIP said:

This overnight closure should be permanent. 

And you're going to put the 120ish trains you just took out of service where, exactly? Politics aside, the fact of the matter is that NYCT was not _designed_ to be closed overnight. Doing so would mean an immense number of mainline layups, whose placement would entail significant disruptions to evening service and cost. I actually doubt that an overnight closure would actually net out to less $$$ than running service. 

2 hours ago, Lex said:

That may be true, but as far as I can tell, frequency is a lower priority for riders than consistency and especially train length, both of which have a far greater impact on capacity and the passenger experience.

The overwhelming majority of research on transit demand shows frequency being hugely more determinative of ridership than the size or shape of transit vehicles. Reliability is indeed important, but the  is a generally well-run route, and its schedule could actually become much more regular if it ran at frequencies closer to those of the ; the 8tph of peak  service has to be slotted in between 11-12tph of , making for uneven scheduled headways. 

On 1/6/2021 at 10:24 PM, Lex said:

From what I've seen, I don't get the impression that  riders are clamoring for a bump in frequency so much as they are for longer trains (sure, more trains would be great, but 6 cars of 75-footers or 8 cars of 60-footers at existing frequencies would be even better, to say nothing of 600' trains).

Because of the >300'/two person crews rule, adding cars comes at a nontrivial cost (at least for weekend service). One important fact to keep in mind here is that  riders have the highest transfer propensity of any non-shuttle route, so wait times between trains matter disproportionately much. I'd strongly suggest starting with service increases because of this (and Court Square can handle 15tph, so not an issue for a while yet). 

I too would love more weekend service, but with current flagging and junction operation practices it's ~impossible to regularly schedule more than 15-16 tph (B div) or 17-18tph (A div) per trunk on weekends. 

Whenever it is that we make some progress on those fronts, I would first bump frequencies on existing services rather than add more. Two infrequent services doing slightly different things is, generally, less helpful from a network perspective than a single, high-qual service.

7 hours ago, Around the Horn said:

CBTC on Queens Blvd was never about ridership. CBTC is now the defacto standard for signalling replacements which is why Culver is getting it.

There are real long term maintenance savings, too. There'd be more if we were less into AWS overlay, but it's still nontrivial.

A lot of the Gun Hilling has to do with congestion around Nereid. Discharging s on the mainline can get nasty, especially if the railroad is already running behind or bunched. It's a tradeoff, but putting those s up the middle and discharging at Gun Hill does prevent delays for riders later on -- whether they be people on northbound s who'd get bogged down waiting for the  to clear, or people waiting for a southbound  late out of the terminal because it got delayed by discharging s. 

1 hour ago, Lawrence St said:

Past Bway Jct in which directions? Towards Rockaway or Towards 8th Av?

Towards Canarsie has dense AWS to support non-equipped mvmts to and from the wash at Canarsie Yard. Towards 8th Avenue has AWS, but the AWS there is low qual -- during Canarise Tube work, a work train north of BWJ => 20 min headway.

(Just to be clear, I do not endorse more AWS. Less AWS is actually better for CBTC reliability and lifetime maintenance costs; we should just equip our work trains)

45 minutes ago, bulk88 said:

NYCT never added TPH to 7, they only lifted a supplement from the 2000s. In any case 7 or L, any trains added are simply avoiding canceled peak trains. QBL pre covid wud be same deal. more E holds at 50 SB Bc E was hot. CBTC QBL is first time I saw the MTA is mixing radio and Legacy trains. 7 and L were forklift cut overs on 1 weekend never going back to fixed block. QBL seems like MTA will never remove the trip cocks and heads. "AWS" forever, double the maintenance parts and double the inspection hours every week.

The supplement that was lifted was one which thinned out the pre-CBTC 27 to 25, IIRC, to support switch work. Net increase was 27 => 29. FWIW, the way the pre-CBTC schedule worked (going back at least to 2003, if not earlier) was alternating 120 and 150 second headways, making for an average headway of 135 seconds, or 27tph. 

I would give it a few weeks/months before coming to any conclusions about the status and direction of the QBL cutover. What's going on at the moment is just that -- a cutover! There have, to my knowledge, not been any changes to the end plan. 

Completely agree with TM5 here. Setting aside Jamaica’s chronic capacity issues and the fact that the yard is almost twice more distant from CRS than CI is from CHU, placing NYCT’s busiest mainline in between a service and its yard is a recipe for ugly schedules and fragile operations. Running stuff up Culver is easy and flexible; QB not so much. 

15 hours ago, Deucey said:

Look at you showing up to say hi...Thought you forgot about us...

So how does this translate for rider experience - are those 6 Av trains moving the same as they do now with 15 second dwell times in station and 45 seconds to the next one, or are they crawling with stop-starts or slow speeds in between stations and 30-45+ second dwells because the distance between trains on non-CBTC trackage will be less than the minimum distance the block signaling allows for?

It’ll be pretty similar to today, just with a few more congestion and merge delays. QB CBTC service plans (pre-COVID, ofc) only projected addition of 1-2tph on each of the  and , which makes for up to 26tph on the segments each shares with the , 25tph on  trackage, and 25tph on . Of course,  sharing will be entirely within CBTC territory, and  presumably will be soon too, which leaves  and  as potential constraints. I think 6th and Culver will be okay with those loads, though, especially now that Church Avenue is working a bit better with the force-and-lock signal mod. 
 

It’s been a busy few months. I’m glad to be back  

3 hours ago, Deucey said:

The thing I was asking is what happens when those QBL trains run on QBL at  or  frequencies in the CBTC zone end up on non-CBTC trackage - 6th Av, Broadway BMT, or 8th Av where there's fixed blocks and interlined services but no wide enough gaps for - example -  to merge at 59th Street and run at the normal speed because three of the next 5 blocks between Lex-59 and 57/7th are red waiting for clearance.

I'm not certain that's what you answered (you gave me academic and tech when I'm an ELI5 guy).

What it really comes down to is the structure of capacity in the system. The combined  and  run more trains per hour than the  and ,  and ,  and  or  and  -- so you get a bump in throughput on the constraining section (the  sharing on Queens Blvd) while the others run closer to their fixed-block capabilities. 

For those who like maps (with pre-COVID capacity #s): 

https://drive.google.com/file/d/1L2j2vVHrd-WJ6yAgkx3Se4sFKLwraDtD/view?usp=sharing

3 minutes ago, EvilMonologue said:

So how would the Eastern Parkway be split up?

It would depend on how much of Nostrand's demand Utica would soak up. You would probably either end up doing  Nostrand/ Utica or  Nostrand  Utica  New Lots, with  >  in tph. 

6 hours ago, CenSin said:

Isn’t Eastern Parkway just as busy though? It could work as a stop-gap measure, but it really needs to be built for B-division equipment so that when an extension to Williamsburg and Manhattan are ready, then it can be disconnected from the IRT (or used for non-revenue moves). We should not be building out more lower-capacity IRT ROWs (like the Flushing extension to Hudson Yards).

Nope. Let's do some math. 

The  today runs 20 trains per hour into Manhattan between 8 and 9AM. For Eastern Parkway:

• The  runs 9tph
• The  runs 8tph
• The  runs 12tph (with another 1tph turning in at Bowling Green)
• The  runs 10tph (with another 1tph turning in at Bowling Green)

The maximum capacity of the  without spending to increase traction power is 22tph; the maximum if you don't spend the untold billions to add tail tracks at 8th Avenue is 26tph. On Eastern Parkway, Lex and 7th Avenue will be capable of 30tph operation after CBTC. Only Lex CBTC is funded at this point, so we'll run with a 26tph capacity on 7th for argumentation's sake (so 13tph for each of the  and ). Looking just at train volumes, we get:

• The  is at 91% of traction power capacity, or 77% of infrastructure capacity
• The  is at 69% of capacity
• The  is at 62% of capacity
• The  is at 80% of capacity
• The  is at 66% of capacity 
• The Eastern Parkway line is at 70% capacity

But train volumes fail to capture how many people are actually aboard those trains. For that we must turn to NYMTC's hub bound data. NYCT defines "at capacity" as being equivalent to 3 sq ft of space/standee. The current NYMTC values for space/rider during the AM peak hour (8-9AM) are:

• 3.9 sq ft/rider on the  
• 5.0 sq ft/rider on the  
• 5.9 sq ft/rider on the  

These values are all measured at these lines' first stops within the Manhattan CBD. NYMTC defines the CBD as being all of Manhattan below 60 St, so the figures we're looking at are for 1 Av, Wall St, and Bowling Green respectively. Given that 2 peak-hour trains arriving at Bowling Green are entering service, it's probably wise if we multiply the  figure by (22/24), or .9167. That yields an average crowding level of 5.4 sq ft/rider.

Putting this in percentage terms, to find the ratio between current loads and the load that would make the line be at capacity (so ridership/(delivered square footage/3.0 sq ft/person)):

• The  is at 78% capacity 
• The  are at 61% capacity
• The  are at 56% capacity

And putting it all together, which really just means multiplying the values immediately above with the track capacity values at top:

• The  is at 71% of traction power capacity, or 60% of infrastructure capacity 
• The  are at 40% capacity
• The  are at 41% capacity

Now, this above analysis probably overstates available capacity on all these lines because their peak load points are not = to the river crossings, but without high precision data on peak loading points and levels, it's difficult to say how bad this gap is. Just something of which we should be cognizant. Even with that equivocation, however, it's pretty clear here that the route with the most available capacity is...Eastern Parkway. It's not even close. 

For our final analysis segment, we've gotta look at some trends. Capacity doesn't exist in a vacuum; a route may have available capacity but if its ridership has been growing strongly over the past x years, it's reasonable to expect that the capacity will be consumed by existing service areas. Below is a chart showing ridership growth at all stations on the Eastern Parkway Line and Canarsie Lines from 2000-present; for all transfer stations on both lines (except for Franklin ) I've split loads equally across corridors serving the stop (so at Barclays, where you have Brighton, 4th and EPW, each would get 1/3). This is an indexed chart, so 2000 = 1 and changes in ridership since 2000 are divided by 2000 ridership to get a multiple of the earlier figure.

...so the only line with a plausible claim of ridership growth constraints is Canarsie. 

The thrust of this analysis should be clear, but if it isn't: the Canarsie Line cannot handle a Utica branch. 

1 hour ago, Armandito said:

There's been talk about a new subway line along Utica Avenue for decades, but rather than create a new trunk line for it, why not just connect it to the in Williamsburg and have it run to the Eighth Avenue terminal? This could be an ideal way to reroute rush-hour short-turns at Myrtle-Wyckoff Avenues, too.

Demand on the  is extremely heavy; adding more people onto already-crowded trains moving through some of the few NYC neighborhoods that are legitimately growing is a recipe for lots of pain down the line. We're better off doing a connection to Eastern Pkwy. 

12 minutes ago, T to Dyre Avenue said:

Would a 57th St crosstown corridor really have parkland as half of of its catchment area? I’m a bit surprised to read that. Even with Central Park just a few blocks to the north, it still seems like a really busy corridor, even post-Covid.

No, I was being dramatic. 57 is probably a bit further north than is best, though; job density in Midtown is strongest between 57 and 34, with the last few blocks approaching the park much more residential. 

....but why, on so many levels?

• Rational human beings within walking distance of the  will walk to those trains and head to Midtown rather than taking a scenic route through LIC
• Rational human beings south of Bedford-Nostrand will probably head to the  
• A nontrivial fraction of the remainder will likely transfer at Lorimer/Broadway/Court Square to services whose Midtown catchment isn't half parkland

Circumferential routes like the  should be permitted to act as circumferentials; I would hope that much had been made clear by the IND's underprovision of Manhattan capacity to the (radial) Queens Boulevard line. If you need to extend the , run it up 21 St, where its utility as a circumferential can be enhanced with the  connection/where it can capture a relatively busy bus corridor. Moreover, I find it...questionable that we're trying to give the  a Manhattan tube before dealing with real system capacity issues like the shortage of subway service in Central/Eastern Queens; that's where tunnels are actually needed. 

7 hours ago, LaGuardia Link N Tra said:

Question for those who proposed the to be rerouted up 21st Street, how exactly would that make deinterlining Queens easier and would it be better for QB Express to go via 53rd or 63rd?

It helps deinterlining because  riders aren't stuck with only 8th Avenue IND service after the dust settles; they retain their 6th Avenue option, as well as access to both QB express and local trains. 

I vote for local to 53/express 63, for all the reasons I've discussed at length in the past. 

1 hour ago, Lawrence St said:

While On the topic about the Franklin Av line, does anyone know why decided to rebuild the line the way that they did, with Park Place and Franklin Av becoming single platforms?

It's an example of intelligent value engineering.  demand is not likely to exceed loads that can be handled on trains running a ten minute headway, so to reduce maintenance costs they built the line's infrastructure to be sufficient for running a 10 with 2 car trains. Neither then nor now has a  to  connection been an at all realistic possibility -- which isn't to say it's not a good idea, merely that it isn't on any list. 

2 hours ago, Armandito said:

So it would've made more sense for the to be permanently cut back to Court Square after the transfer to the and at 23rd Street-Ely Avenue was opened in 1989. And if you're saying trains were mostly empty after Queens Plaza, it's most likely the to Forest Hills was kept purely for political reasons. To make an analogy, the to Forest Hills was basically the old to Coney Island via Brighton before 1986, and then to Bay Parkway via West End from there till 2010. From what I read before, the was not useful as a part-time service anywhere in south Brooklyn.

Well, not exactly. Before 63 was linked with QBL in 2001, your Manhattan capacity was limited to the , and the  alone was not enough to sate demand on the local tracks. That’s the role the  played.

As for the , the demand demographics along its route were actually quite downtown-biased relative to the rest of the city until the aughts. Then that changed, we needed to cut service, and we got the . 

2 hours ago, Armandito said:

Understood. Now suppose the never came to be and the still operated to Forest Hills. Though it may still be possible to turn X trains at Court Square by relaying on the middle track in this scenario (the same track trains use to relay there), it would seem rather unusual because terminating trains would have to go the opposite direction to access the track and then return to the other side of the platform to depart. This would be the only way to turn there if there still were to be through service to QBL, assuming the alternative scenario I envisioned here.

The 63 St Tunnel exists now. Absent a massive reorientation in NYC job geographies, there simply isn’t any reason for us to be wasting QB capacity on the , so it’s really okay if we build infrastructure that impedes that operation. 

2 hours ago, Armandito said:

@RR503 How's this track layout? I removed the second platform and introduced a scissor crossover this time:

I don’t think you need the layup tracks (the operational benefit of being able to put in trains from/stash trains at Court Sq vs the cost of underpinning a line just doesn’t pan out in my eyes), but otherwise this is what I’d propose for the area should we ever need high throughput terminal ops there.

23 minutes ago, Armandito said:

As in if the flyover were to already be built and in existence?

Yes. But again, I don't think the flyover is even a rough approximation of a good use of funds.

8 minutes ago, bobtehpanda said:

If any extension of the is to be made the most logical one is probably up 21 St.

Which has the nice effect of allowing you to make a connection with the  at 21-QB. Makes deinterlining much easier. 

7 hours ago, Armandito said:

IIRC, we had another member here propose a / terminal swap by building a flyover somewhere between Neck Road and Sheepshead Bay so more trains can terminate at Brighton while simultaneously enabling through express service between Coney Island and Manhattan by way of the . In this case, locals would end at Brighton while expresses would continue toward Stillwell on weekdays, and at other times when the does not run, the is extended to and from Stillwell. With this kind of layout, the two outer stub tracks could be used to store additional trains (thus allowing for X's to terminate here as a result of the slight increase in terminal capacity).

As I said when that proposal was first made, such a flyover would be an extremely low-value investment benefiting riders at only two stops. Really don't think it should be made. And if it was made, you could easily operate Brighton Beach as a relay, which would give you plenty of capacity to terminate  and (X) on the express tracks. 

7 hours ago, Armandito said:

Regarding Court Square, I proposed reconfiguration of the existing tracks and the building of a new platform so the station could be used as both a terminal and a through stop. Another reason is this: during a few trips from that station on the recently, I would see that both tracks would be occupied by turning trains. If you say that a scissors crossover would seem better, I would suggest building layup tracks north of that station instead--in a manner more or less like the layout at the Broad Street  station in lower Manhattan.

There's no need for Court Square to be both a through and terminal stop. Absent the construction of a new line (which hopefully would be valuable enough to be served by both the  and (X)), any extension of the  beyond CRS constitutes a negative-value proposition; you're reducing Manhattan-bound capacity on the Queens corridor. 

As for why there are trains in both pockets...schedules! If you're running a low throughput terminal, you're going to have long terminal layovers, and periods where both pockets are occupied. If you added service, you'd simply shorten layovers (the  averages 2-3 mins at 34H) and go from there. 

You really should just adjust the track layout there so you can cross from the local to the spurs and then onwards to the express immediately after Brighton Beach. You’d need to move the x between the expresses south a bit, but that’s trivial in the grand scheme of things. 
 

Neither relaying or terminating at Ocean Parkway is possible without significant alterations to the interlocking logic there. Having spent many an evening watching  layups at the location, I can tell you that when the switch is lined against you leaving OP (which it always would be), you cannot fully berth which is a dealbreaker. It’s probably tractable, but get ready to shell out a couple mil for a bunch of new IJs and ST-equipped signals.

Re: Court Square, we turn 30tph at a 2 track terminal at Hudson Yards. If you modify the track layout south of the station at Court Square to match (ie replace that spur with a simple scissors crossover), you should have no issue turning the  and (X) without spending millions or billions on an annex.