JamesRR wrote:My NEC train was over 10 minutes late this morning. It hadn't been late all summer - but it returned to doing its usual 'crawl' from Secaucus to Penn. Ultimately the delays were blamed on "delayed trains ahead."
The reason why your train was on time all summer is simple -- there were fewer trains on the rails between Newark and NYP. With fewer trains there was some 'idle' time for the train traffic to recover from random delays and such. Now that we are back to the usual number of trains, we are back to the usual state of delays.
Amtrak will tell you that they have a high capacity signal system installed between Newark and NYP. Don't believe them. No, wait, I am wrong, maybe, kind of, for your purposes as a commuter I am correct ... What they have is a signal system that can support more trains per hour on two tracks than is installed anywhere in the US. It is a better system, true, but it is taxed to the limits. The schedule, especially the morning commute schedule is so filled up with trains to the physical limits of what is possible with the two tracks and the signal system. The problem being at the limit is that any one delay propagates and causes cascading delays because there is no 'idle' time allowing for recovery.
A pure two track railroad can easily support more trains per hour than what we have now going into NYP, easily another 10-12. The problem is that we do not have a pure two track railroad: Kearny Junction and Secaucus Junction Station mess up the pure unidirectional flow on each track. A pure two track railroad between Newark and NYP will for the most part be limited only by the ability of NYP to platform trains (generally the approach to NYP is the slowest part of the segment). Secaucus throws one wrench in this by -- primarily because of expresses that do not stop there overtaking locals that stop there. Imagine a sequence of express-local-express coming from west of Secaucus. The first express passes, the local stops, the second express overtakes it. What do you get at NYP? Express-express-local. It is not that simple though. What you get is express-nothing-express-local. The time slot between the two expresses for coming into NYP is 'lost'. Kearny Junction throws a similar wrench in where you need to have a gap between two trains coming from Newark in order to merge a train from Newark Broad in between them. These create a very complex dance when many trains are trying to get to NYP early in the morning. If one train merging from the Lackawanna lines is late or if due to heavy travel or delayed Erie trains a certain Secaucus Upper Level train takes an extra minute to board you easily get the other trains behind delayed. And it propagates from there with little chance of recovery because the schedule is set to the maximum train density possible with the existing hardware. It can only get worse from there with equipment breakdowns or with Portal Bridge openings.
There are only two ways for the on time performance to become better. One is to cut the number of trains -- you already experienced this. The other is to provide more tracks on at least a portion of the segment between Newark and NYP. If you want the on time performance to improve you better hope that Congress and Trump provide the money for the new tubes under the Hudson -- and even then, the improvement is about a decade away. After the old tubes are rehabbed, there will be four tracks between Secaucus and NYP. For even better results, you should hope for quad-tracking between Secaucus and at least Kearny Junction (better to Newark, but the effect on on time performance of the segment between Kearny Junction and Newark is likely minimal). Till then there is not much more to do than to enjoy the delays -- bring a book or watch a movie on your iPhone.