RR503

WTC honestly...isn't that bad. It's a stub, but you could (esp. post-CBTC) probably get >>20tph out of it, if we use 8 Av and SF as our reference points. The issue with interlining 36 St is twofold. The first problem is that it eliminates any incentive to stay on the express beyond Roosevelt -- riders can use the and riders can use the . The worked largely by giving ridership beyond Roosevelt a one seat ride across the peak load points into the CBD (thus relieving the ) and picking up nontrivial LIC-6th Ave ridership that used to use the . Very, very few people through rode beyond Roosevelt when the express was an alt for their destination. Second problem is, of course, ops. I don't think I need to explain just how garbage 59 St is. Its peak throughput is about 43tph across two tracks. Imagine running it at 50-60tph. It's simply an operational non-starter.

Something to be thankful for here is that the T/O got the train to 110. Tunnel fires are scary; tunnel fires in deep bored tunnels built to 1900s egress standards and with little fire protection could have been even more catastrophic. My heart goes out to the family of the deceased; he's a hero.

Gonna disagree that it's a great idea in theory -- it's a great one in practice as well. It was just designed obtusely, so that you have to LR the entire train vs a zone, which means you can't use it in normal door operation or outside of an autorecycle sequence bc just pressing the button may recycle a door you're not looking at.

When the signal contract went out for Culver, Church was never supposed to be a real terminal -- it was just a place to squirrel away trains. When presented with the opportunity to save maintenance $$$ on 4 switches, then, the agency went for it....et voila. I'm not either. But it's certainly something worth looking at IMO.

Church certainly could use a little help on the policy side, but there are real infrastructure constraints there -- especially if NYCT ever lengthens Gs to >300'. The ramp down to the lower level is timed to 10mph, making the diverging move down to the layups S L O W. The ramps themselves are also quite short, so if you get downstairs and don't have a lineup into a relay, a long train will overhang onto the main. Once you're at the yard, things don't improve. Some aspiring engineer decided to mess with the switch config in the yard durin the resignalling, so instead of having this: We now have this: Which reduces flexibility and forces the installation of annoyingly restrictive signalling. The relays themselves are also only 600' long and have AK signals on them, so if you're a long train you're gonna c r e e p in, reducing turning capacity. These problems, with the exceptions of the ramp and relay length are fixable, but would require a decent bit of investment. When the alternative is good for ops, good for the budget (because of the amount of time it takes to relay a train, doing the 8m + layover time trip to 18th would be approximately cost neutral but ridership-positive), and good politics....why not? The 4th Avenue corridor has 3 branches and 2 tracks in each direction. You're stuck with a merge kinda however you want to slice the pie, so to speak. The move off of West End is nasty, so I'd imagine that new xovers south of 36 could only be an improvement. Free yard space! Also overflow for whatever you can't turn at 18. If you move the crossovers, you could definitely do more than 15tph. Both levels have tail tracks (albeit ones that are normally occupied), so with good switch geometry I don't see any reason why we shouldn't be looking at capacity figures that begin with a 2 or a 3.

Ah, sorry--CBTC can fix all of these areas, but they're only installing on 8th and Fulton.

See my original response. CBTC fixes this 100% because your speed profile is the lower of maximum allowed speed for geometry and maximum safe speed for following. This, plus the acceleration benefits that come with CBTC ops are why you see such big runtime gains when you turn CBTC on. The whole point of my reply is these aren’t exactly physics (curve speeds etc) problems: they are choices made relative to a very specific set of constraints inherent to fixed block installations. FWIW, CBTC fixes operator variability but not conductor variability, which can be punishing in high ridership segments.

If I were a betting man, I would be willing to wager that a majority of timers exist for control line safety (ie making sure trains are going slow enough that, given a certain distance ahead of a signal which makes it red/maximum attainable speed, there's sufficient stopping distance) rather than to protect nutty track geometry. Most CPW timers are there for that reason, or were baked into the resignalling they did there in the late '80s because controlling train speeds actually can help _increase_ capacity (this is the primary motivation behind many timed areas on downgrades: it's not so much that we don't want trains doing 65 as it is we don't want to have the control line lengths that'd come with that).

Yes, West End loves express. But this gets you yard access for the and doubles their frequency or allows express. I'd say that's a viable trade-off, no? Because Stillwell can't handle that many s. Absent some rebuild plan, you need KH as a relief for s, which forces the to 18 or Church. Of those two, I'd take 18th with the equivocation that at tph >10 you should split between the two terminals as you really don't want to be running a single pocket at those frequencies. Ah, I see. That makes sense, though you could easily use Crescent or 111 St to short turn. Much less pretty of an operation, but doable enough. I see on your map that you moved the crossovers on the UL at Parsons -- why not do the LL too?

CBTC will take care of 8th and Fulton, but you can't really do anything about the timers on CPW. Whatever you may think of their implementation, they do indeed serve real safety functions!

They stopped turning 10s into 8s or 6s for that reason. Splitting a train in half isn't nearly as annoying an operation as doing some uneven cut, as you end up with some stump when you're done. Lefferts , Dyre , Myrtle do this every night. Shortened trainsets would free up CRs and reduce ABDs related to CR availability, but you wouldn't save that much $$$. Remember, a work program is a work program -- whether you run what's in it or not, those called on a given day still have to show up.

So I guess my question is why you chose not to do things like to 18 and to West End? Those things have clear operational (and ridership) upsides, which have been talked through in the past/in this current discussion. There certainly are counterarguments to be made, but "I didn't feel like it" isn't really one of them. Also am wondering what your reasoning with the is. At 16tph, the schedules for express allow a merge at Bway Jct without too much pain, and there's nonzero demand between Bway Jct and points east. Why short turn locals?

On the short term map, express/ local allows you to deinterline 59, 50 and Canal without touching Queens and helps keep the to a decent length. Once you've done that, you're somewhat forced to do to Brighton -- to 4th would mean you either end up with 4 weekend services on CPW (the two expresses, and then two locals because you need one for each branch of 4th exp), more local than express service on CPW, or one of West End/Sea Beach without weekend service to Manhattan. On the long term, I figure a) continuity is good, b) the logic rel. the still holds, c) the as an overlay Brighton Express/Concourse Local service is legible and really pretty to operate (clear hierarchies of primary/secondary services makes disruption management easier because it's easy to thin out/suspend a train without messing up some branch's service).

Honestly I don’t know. It’s unclear whether this issue is a function of the way NYCT zone controllers/interlockings interact with CBTC, whether this is some AWS overlay bug or something else entirely (perhaps related to our conservative design assumptions?).

It's sad how the transformation plan frames the choice between reform and the status quo one between a poorly thought out RIF and not doing that. The root of ~70% of the MTA's problems lie in process: fix those, and you'll realize huge efficiencies, and will be able to do RIF (which is needed) without affecting vital operations. This will be interesting to watch, both in the near term w/ Corona, and also thinking long run, because a ton of out-year budgets are predicated on these Transformation savings...

What UT said. The only equivocation I'd make rel. CBTC and capacity is that current NYCT CBTC architecture locks trains out of stations until the train in front is fully clear of the platform. Doesn't make a huge difference on stations where the leader can exit quickly, but at Cordlandt and City Hall (especially City Hall, given entrance speeds there are also slow) you'll need that ability to achieve full capacity.

Yeah, they had some turn there when Montague was out.

If it doesn't, we're likely going to have a front-line seat to a wave of transit authority defaults/bankruptcies. There really isn't a way for the agency to cut its way out of this -- the gap is too large. This isn't to say that I at all disagree with your assessment here, just saying this will be catastrophic.

Maybe a 3-4 per hour? But even then isn't good -- the second you start turning trains there, you're going to cause irregular gaps in through-running service unless you're efficient. Which they aren't. There's aspirational and then there's ~aspirational~ though, and given that I'm not convinced of that project's relative merit... A stub end terminal with meh crossover placement? With CBTC, idk, 18tph? Which is more than you'd ever need. Would be nice, but that means grade separating the junction. Don't think that's an easily justifiable investment. Up until spring 2018, there were s that entered service at Canal/Tunnel at 7:55 and 17:24. I don't _believe_ we schedule trains to lay up there outside supplements anymore (though don't quote me on that, I'm just skimming schedules and there may be deadheads from DIT), but it's certainly a known move for the agency. In years past, you're right that there were even more such trains -- in 1991, 4 s and some number of s entered service there for the AM rush.

Looks like they had to ABD one interval, and then just ran the rest as s.

It wouldn't save much money. All the weekday crews would still have to report, which would mean you wouldn't have to dole out O/T to cover missing crews' assignments, but that you'd still be paying the lion's share of your normal cost structure. Something to consider, but it really isn't perfect either. Writing schedules is an extremely complicated process. Your average supplement (what's used to run diversions) takes months to write and coordinate; the process for assembling the base schedule used in the pick starts something like 18 months out. Even if you did pull together a supplement, though, you'd probably not even end up saving all that much. The base work program is an optimized fabric that minimizes the number of crews required to run service (with some flaws, but this is generally true). As soon as you start picking at that -- even if you're running less service -- the number of crews required actually usually goes up because that fabric gets disrupted. This is why GOs that put the on a 12 or shortline the to 145 almost invariably require more resources than their base equivalent.

You need to preserve rail up there for yard access — you're going to need that put in capacity if you want to deinterline Nostrand without blowing your ops budget. If you’re going to have rail up there, you may as well have a shuttle. It’s not perfect by any standard, but it’s politically better than nothing, and should be relatively cheap to operate. A mistake — cut the flyovers. As you can tell, this went through many iterations before I was happy with it; originally I had express and local, but decided that it’d likely be mighty difficult to justify such a significant investment when a zero cost alternative that had the same operational impact existed. I’m very open to arguments the other way, but that was my reasoning.

You'd rebuild 135 -- move the s/b track to where the spur is, move the platform with it, and add a 3rd track alongside the western edge of the 135 sb plat for the shuttle. You'd have to build a bit of tunnel north to 142, but that should (?) be doable?

You could, but again, I'm aiming for investments that could concievably get done by the MTA in this day and age. I also am still somewhat unconvinced that this specific project should be a priority -- it'd be $$$ for a nontrivial operations gain, but only a small capacity gain. If you reconfigured it a bit (ie abandoned 4 tk and used M and 3 only) you could probably extend the western island. Then you'd just need to signal it properly... It's something I'd like to see looked at, though only if it's found that CBTC cannot fix the curve issue.

Maybe not 30, given that Concourse and CPW aren't _that_ high ridership, and that returns to additional trains really start to go down >20. 24, say? And yes, I'll add those.