djimpact1 wrote:Aside from the vehicle procurement process/selection (to which I understand the T's on-and-off history of being successful at it), what necessarily is bad about DMU implementation?
My personal opinion is have DMUs running weekdays (exclusively) during off-peak service for the 3 lowest-ridership lines: Fairmount, Greenbush & Kingston/Plymouth. Concerning weekends, DMUs would run exclusively during all hours of service for all lines that don't see at least 2,000 riders per every Saturday & Sunday: Needham, Fairmount, Greenbush, Middleborough, Franklin, etc.
Why bother using the same resources for peak ridership (a diesel engine, a single & a few bi-levels) as for non-peak & weekend ridership, especially considering the numbers can be drastically different? If you know a line might see 500 - 700 riders in an entire weekend, it seems foolish to operate a full set that I can otherwise understand seeing during weekday commutes. At least DMUs seem like they'd save some miles for the "big fleet" needed for those high ridership lines/days, while still giving a ridership option to folks looking for that off-peak commute to Boston.
Thoughts? (F-Line, go easy on me brother!)
It doesn't quite work that way.
Push-pull starts to overtake DMU's on performance the longer a route gets and the wider-spaced the stops are. The difference is in the diesel engines: DMU's have 2 high-performance engines, diesel locomotives a single low-performance engine. On a regular loco the electricity for the traction motors that drive the wheels is generated by a big main engine with large surface area that spins at (relatively speaking) slow RPM's. High-performance engines shrink themselves to small size by spinning a lot faster. In addition to DMU's, true dual mode locomotives like NJ Transit's ALP-45DP use high-performance diesel engines to shrink the engine enough to fit a full electric loco's worth of guts inside the carbody. And DMU's need two of these engines because at the very very small size those engines have to be to fit underneath the carbody they don't generate enough individual power to go it alone except when one engine fails and the other can keep the train going at restricted speed.
"Low" performance has a couple advantages. By spinning slow at a large surface area they do do much better fuel efficiency and emissions than high-performance engines relative to the power generated. And because they generate more power at slow RPM's they're very very efficient when they're either idling or cruising at full track speed. It's how those CSX commercials make their claim "we haul X tons Y hundred miles on a single gallon of fuel". The trade-off is that they have to work hard on starts/stops to carry all that bulk into motion. High-performance engines, because they spin fast, are less fuel-efficient at idle or track speed. NJT's dual-modes are guzzlers compared to similarly modern diesels of equal power. And DMU's, despite the much-reduced power required by each engine to push the trainset, are less efficient at full-speed or idling on the platform than a much bigger push-pull. Their primary advantage is starts/stops...not having to push so much weight into motion means less of a differential for how much the engines have to rev up on acceleration. As you lash up additional DMU's into a consist, you're burning a lot more fuel at cruise and idle but also having a much easier time accelerating from a stop.
It's a trade-off, and you can see where each has a performance advantage and where those advantages start to converge to evenness and eventually trade places in overall efficiency at a certain total distance and a certain decreasing stop spacing. An HSP-46 pulling a single car off Track 61 to West on a less-than 5 mile, 3-station, slow trip through the terminal district is going to be a pig to operate. By the same token lashing up 3 married-pair DMU's to run the whole 63 miles to Wickford Jct. on a Sunday is going to be a pig because the route's long, the station spacing is so wide it's significant stretches of 79 MPH (or 90 MPH) cruising, the consist has 6 diesel engines, and it's going to need more frequent refueling because the fuel tanks are small in addition to the engines draining the tank faster than the much bigger loco (i.e. it's going to be out-of-service more service hours in a given week in line for refueling if it can't make nearly as many trips in between refuelings). Use each vehicle type to their strengths or the operating costs start to bloat across the board. It can go wrong with either one.
DMU strengths -- Inside-128 lines that run 10-15 miles. Lines that have stops every 1/2 to 1-1/2 miles. Lines that have relatively short (non-terminal) platform dwell times, such as quick-hit interzone trips. Lines that don't require a lot of cars or require huge swings during the service day in number of cars. Usage that lets them cover higher number of runs between refuelings. Usage that lets them return immediately inbound with without long layovers (i.e. closer proximity to terminal-serving yards like BET and Readville instead of having to spend much time at the outer suburban layovers).
Push-pull strengths -- Outside-128 lines that run 20-50 miles (Wickford and Fitchburg are both 50+; Worcester if 44; Newburyport is 36; Kingston, Middleboro, and Rockport are 35; Haverhill is 32; Forge Park is 30; Lowell is 26). Lines that have 2-5 mile stop spacing with ample stretches of full track speed in between, lines with crowd-swallower stops where extra minutes of idling needed to load/unload an overstuffed platform. Lines on freight-clearance routes that have to have a lot of stops with low platforms + 1-car mini-highs instead of full-highs, because those platforms take longer to load/unload. Lines that require high seating capacity (loco engine gets progressively more efficient with each additional car), or have very large and distorted peak/off-peak swings in capacity.
DMU weaknesses -- Long lines. Wide stop spacing where the engines work harder to maintain track speed. Lines with long dwells. Lines that need > 4 single-level coaches' (i.e. >2 married pairs) worth of seating capacity to run the schedule. Clearance routes...depending on line (the 1-car mini-highs alleviate the problem but they're positioned differently at each station and sometimes that'll make the unloading spot awkwardly-placed...highly situational stop-by-stop). Lines that increase the frequency of downtime for refueling. Schedules that require longer outer layovers instead of immediate turnarounds (more fuel burned at idle, layover facilities may need modifications to their locomotive plug-in pads for overnight storage of DMU's). Lines that cross staffing districts requiring greater pool of certified engineers, conductors, inspectors, etc. qualified on every vehicle type (e.g. outer layover staff will need to be trained on inspecting vehicles for problems, Rhode Island has its own crew base that has to be brought in).
Push-pull weaknesses - Extremely frequent starts/stops burn fuel. Slower acceleration at tight stop spacing (note: makes more difference on straighter/higher track-speed lines like Fairmount than slow/curvy lines like Track 61 or Grand Junction that'll never get much above 30 MPH even with perfectly maintained track). Minimum consist length currently 4 cars; operating/maint regs have to be loosened to allow 3 or fewer like Metro North and Amtrak allow. While probably doesn't require many more locos or blind coaches to supply such short/quick runs, may need to get more cab cars for these short consists...and the only spares are 25 of the 1600-series Bombardier cars that would need their deactivated controls and signaling equipment reinstalled.
So...you can see why this is a bad idea for the 495-and-beyond lines any hour of the day. The operating costs do not wash despite the lower seating capacity. Remember, an HSP-46 is designed to haul 8 rush hour bi-level cars packed standing-room only to up to a 90 MPH track speed. 4 cars with the last 2 cars blocked off so they can staff 1 fewer conductor is such a light load the engine's not working very hard at all doing Providence or Forge Park late on a Sunday night. A married-pair DMU would be working harder. That's something they'd never want to do. The costs would wash worse. What they need to start doing is putting together more precisely-sized consists. MBCR really was haphazard about that, and that's why you saw so many overstaffed 5-car trains when 4 was fine. Keolis will probably be better at that simply by not flaunting how little they care. Also, those 5-cars don't have to be 5-cars all the time with the fleet now going bi-level. What ridership spikes do happen on at certain slots on the weekend they don't have to hedge so much at going long and can stick more conservatively at 4 cars with their margin of error being the bi-levels, not the extra car. Likewise, improved staffing levels can help. The 1 conductor per 2 cars rule works a lot more efficiently with bi-levels instead of singles, and even-numbered consists instead of odd. So put together 4 cars on the weekend, not 5, and spare the extra crewmember. And have a quicker hook at closing unused cars and doing front 2 cars only when a long (5-8 car) consist is on one of the first/last empty off-peaks before a peak period, or is assigned to a late-night slot for purposes of getting it to the outer layover for Monday morning's rush. All of those efficiency-minded practices will cost less than weekend DMU's systemwide.
Have to understand how unusual the RDC's-everywhere era was. Boston & Maine was the only passenger carrier in North America that replaced steam with 100% DMU. Nobody else did that. While they had fleet standardization controlling their costs, they had the same forced one-size-fits-all efficiency limitations that a DMU revival would have. LIRR, NY Central, and Pennsylvania RR primarily deployed RDC's on specialty services like outside transfers into the electric district (same as LIRR's diesel transfer shuttles today). CN and CP in Canada bought also primarily for specialty reasons: to serve the most remote, car-inaccessible settlements in the country they were mandated by the government to provide passenger and mail service to. NYNH&H was kind of #2 on the weirdness scale for how much the RDC's appeared on the South Station branchlines, but even they didn't go nearly as whole-hog as B&M did.
While it was a very progressive move on B&M's part and arguably kept some extremely marginal one-per-day branches barely alive enough to pass into public ownership (and in some cases survive to this day)...it arguably wasn't the greatest business decision on their part at a time when they were making a lot of less-than-awesome business decisions. For the same reasons the T going overboard trying to shoehorn them into the wrong situations or not running them often enough in the right situations wouldn't be a great business decision. Most railroads across the country went straight from steam to diesel loco, then immediately put their investment into cab cars to transition from pull-only to push-pull ops. And invested in a lot of general-purpose power that could be geared for passenger and freight. We weren't necessarily the vanguard up here in the New England...DMU commuter ops were the oddballs. Oddballs whose usefulness came later when they helped keep commuter rail breathing with a weak pulse past the bankruptcy era into the public ownership and rebuilding era. You can't chalk DMU's up to a revolution deferred...it wasn't like that the first time around either. Loco-hauled was always the every-vehicle, DMU always the specialty vehicle. Use them where they work best, and mix both instead of trying to shoehorn specialty vehicles as every-vehicles.