BandA wrote:Intermittent occupancy tracking? That seems unsafe.
It's not unusual at all. Old pole line systems still extant on a shrinking portion of the northside don't always transmit continuously through the running rails, either. You have isolated track circuits at the block boundaries same places where the wayside signal heads are plugged into the pole line. Some really old ABS, like what used to be run on the Fitchburg Line out of Waltham Tower, is so primitive it can't do any train detection at all and relies entirely on regular human-to-human radio squawk to ascertain position and make dispatcher decisions. None of it is inherently unsafe because of the great precautions taken through the rulebook. But non-constant tracking comes with obvious tradeoffs for MAS and how densely you can pack trains on the line. Fitchburg's schedule used to be so irredeemably awful less
because of the S. Acton-Willows single track than the fact that the whole of Somerville-Ayer was bound-and-gagged by signaling techniques virtually unchanged from 1910...and all margin for error was chewed up and gone before having to stage meets around the single iron mid-line.
The drag effect on the inner Worcester Line is self-apparent. 59 MPH MAS where the track and geometry are plenty good for 79 MPH and would easily hit 79 inside Route 128 on the schedules that skip Newton. Lots of fudge factor required in the schedule because of the lack of tracking precision, giving the line its excruciatingly long end-to-end travel times through excess padding and contributing to its infamous near-daily cascading delays when one late train fouls the margin-for-error for everything behind it and every planned meet inside-128. And the max achievable train spacing with that inflexible block/detection layout leaves an artificially low ceiling for service density, inhibiting much additional schedule expansion. Low enough that the addition of CP6 crossovers and fixing the Newton single-platform pinches are no-way/no-how enough by their lonesome to serve up slots for layering Indigo-Riverside service...and probably not even enough by their lonesome to serve conventional Worcester service growth beyond another 5-8 years of incremental schedule backfill. This despite Worcester being the least
capacity-constrained by South Station terminal of any southside line, owing to its conflict-free platform assignments on the far Atlantic Ave. side. And this despite Worcester being the only southside schedule besides Fairmount that doesn't have to play branching games on the NEC or amongst itself.
Continuous track circuits the likes of which you find on every wayside-only line (excluding pole line ABS Rockport and Western Route on the Wyoming-Reading double-track) fix those handicaps through gapless tracking. The cab signal layer--which bootstraps on top of the bottom-layer continuous track circuits--adds the speed regulation. And then the ACSES PTC layer on top works in conjunction with the cab signals to slam shut the last one-in-a-million safety loopholes and--if programmed to-task--fine-tunes the train spacing to dynamically pack 'em as close as whatever desired tradeoff of speed vs. density/meets is optimal for the layer cake of services they want to run. But it all starts at the base, which is why the signal system has to be cleanroomed to punch through the service ceiling currently imposed. ACSES can/will operate without cab signals and be able to backfill the functions of that missing second layer...albeit at more up-front cost and complexity and fewer recourses for avoiding draconian stops and speed restrictions in event of a chance signal dropout. But since ACSES is only designed to augment
the lower signaling layers, not supplant
them...the achievable service density is still limited by the lowest common denominator. Dispatch gets a little bit more computer-brain assistance on decision-making for managing those cascading delays, but there still aren't avenues for implementing "layer-cake" exponential service increases or stamping out the existence
of the cascading-delay vulnerability inside-Framingham without the $$$ of a full-on replacement. Meaning, no perma-fix...no Indigos...no overlapping service layer cake until you trench 20 miles of brand new fiber for continuous track circuits, cab signals, and the type of cabbed ACSES the NEC has and Old Colony main will quickly have for threading their much greater overall schedule loads.