Deinterlining: A thread.
So before Covid @A320Lga and I were working on a detailed report showing how interlining limits the subway capacity. One day the full report will be released but for now here are some highlights to show just what is possible with minimal investment!

Interlining, or branches running together along a trunk line, isn't a terrible idea. But NYC takes it one step further and has reverse branching where now 2 or more trunk lines are intertwined. A delay on one branch now can affect 2 or more trunks, themselves interlined.

Because of the nature of this complex system some tunnels run under capacity. We found that as a whole, pre-Covid, the subway ran only at about 60% capacity in the AM rush (the busiest time.) Some lines are obviously busier than others.

I've written about fixing interlining in Brooklyn through DeKalb before. The quick and easy visual is below. Segregate the N/Q to the Brighton Beach Line and B/D to 4 Av Exp. With a new switch south of 36 St you can provide more/exp service to Bay Ridge.

Sending the R along the West End Line (currently D) allows you to run up to 21tph there which is enough service to make up for the loss of direct express. Total trip time with the reduced wait is the same to Atlantic Ave. (For more detail on how https://www.vanshnookenraggen.com/_index/2020/10/deinterlining-with-one-switch/)

On the other end you need to fix up Astoria. With the addition of new switches south of Astoria Blvd station and reconfiguring the interlocking into Ditmars you can turn more trains at both stations. TPH on the Broadway Lcl track is 21tph, limited by the sharp City Hall curve.

Queens Blvd can be reconfigured with no extra work. Take the R away and send all F/M trains via 63rd Tunnel and run local to Forest Hills (Jamaica at rush for extra service). E (Parsons) and a new K (179) train take over express via 53rd Tunnel.

I'm open to the idea of having F/M as express and E/K local but the problem is M trains run with shorter cars. Not good for how packed the express trains get. Extending the M platforms in Brooklyn is a great idea, it just adds to the cost.

8 Av and CPW would be changed: A lcl from 168 to WTC, C lcl from Bedford Pk Blvd to WTC at Rush. B exp from 207 to 59 St, D unchanged. This allows the E/K to run exclusively on the 8 Av exp tracks and on into Brooklyn via Fulton St.

Queens Blvd express needs much more service than CPW. The WTC terminal can't handle the extra service that CBTC will bring so the smart move is to just send all that traffic to Brooklyn and use WTC for the lower levels on the A/C.

On Fulton St E trains run express out to Rockaway, both Far and Park. K trains take over the local tracks to Euclid Ave and possibly out to Lefferts Blvd (although this adds another merge). Interestingly the E from Jamaica to Far Rock is slightly shorter than current A train.

The IRT is where you need to spend the big bucks. Everyone knows about Rogers Junction which shouldn't cost that much to fix (yet the MTA will find a way...) But in order to really get the most out of the fix you need to build layup tracks south of Flatbush Ave.

This fix would see 2/3 trains running to Flatbush Ave and 4/5 trains to New Lots. I've heard rumblings from local pols who don't want to see the loss of a OSR to the east side but with higher levels of service and fewer delays a cross platform transfer is quicker.

Harlem and the Bronx need love too. The 149 St merge is just as bad. Forces 5 trains to crawl through the curve. A nasty at grade junction on the 2/3. The solution seems simple enough until you realize that simple deinterlining screws up a lot of commutes.

Harlem-148 St is a very popular station line so you need to convert the section into a shuttle. This can be done pretty simply by adding a third track on the east side of Lenox Ave and expanding the station.

149 St is tricky because Lex service is very popular on the White Plains Rd Line. To deinterline it will be important to make transfers between the 4/5 and 2/3 better. Expanding mezzanines and adding new egress points will help. A new Metro North station should also be considered

Sending the 5 up on the Jerome Ave Line will require reconfiguring the interlocking at Bedford Park Blvd so trains can enter/exit the yard without impacting 4 trains running to Woodlawn.

If I ever get around to releasing the final report you can see our work. One thing I haven't worked out is the cost. The switches should be well within the maintenance budget of the NYCT unless the MTA wants to grease some palms. The bigger builds will need to be studied.

A big omission here is the Jamaica and Myrtle Lines. That's another big build that requires its own thread/post. For now these are the "low hanging fruit" proposals that could get us to seeing 2-3 min headways, or as @scottmstringer @NYCComptroller puts it, "NYC in 6 min"

There are more fixes possible. These are the big ones.
A big thanks to @alon_levy for their original musings on how to deinterline.

@RidersAlliance @Union_Tpke @PCACriders @NYCTSubway @mta