mtuandrew wrote:Someone at Amtrak would have to really love the GEVO to want it in an existing Genesis. The Gennies' 7FDLs are fine for 4200 hp and can reach a high enough emissions tier for rebuild standards; it'd be foolish not to examine rebuild options, especially given the seeming success of NS Altoona and their AC44C6M rebuilds.
The AC44C6M's aren't comparable because they're rebuilds of Dash 9 series engines, while the Genesis is a Dash 8 series. It's well-established that the microprocessor controls are the one Genesis component most suffering for parts supply and the biggest unsolved variable for estimating rebuild costs. That is directly tied to their lineage in the Dash 8 series.
7FDL prime movers were featured in the Dash 7 (mid-70's to mid-80's), Dash 8 (mid-80's to mid-90's), and Dash 9 (mid-90's to mid-00's) series. Dash 7's had analog circuitry in their modular electrical cabinets, much like the EMD "Dash 2" classics using the 645 prime mover (GP38/39/40-2, SD38/40/45-2, F40PH, etc.). Dash 8's introduced digital microprocessor controls for the first time using circuit boards in the electrical cabinets. Dash 9's were second-generation microprocessor controls that transitioned more of the circuitry from custom hardware chips to microcode and programmable firmware. Today's third-generation microprocessor controls in the GEVO's and competitors pretty much completed that transition from hard-printed circuit controls to reprogrammable firmware, and are now pretty much "more software than hardware".
Dash 7's, like their EMD Dash 2 competitors, are very easy to upgrade because there's no legacy computer brain to tend to. Many have had their electrical cabinets retrofitted with new microprocessor controls to aid in upgrades like new emissions controls. The Dash 9's, being the closest generation removed to modern firmware-driven power, are getting rebuilt en masse by Class I's because their second-generation electronics were built with an upgrade path in mind.
You don't, however, see as many Dash 8's getting rebuilt above-and-beyond their original factory specs...or beyond an uprating/downrating from one factory spec to another (for example, the P40 remans bringing the performance in line with the P42). That's because those more primitive first-generation microprocessor controls had all-hardware logic boards custom-produced for each make, and none of the reprogrammability of the later generations that relied more on firmware. The upgrade path stopped being easy about 10-12 years ago when GE started curtailing direct service & support for the older lineups and their hard-to-produce hardware boards and concentrated its energies on the firmware-heavier Evolution and Dash 9's. So to outright upgrade a Dash 8 requires third-party replacement or reverse-engineering into firmware of a whole lot of custom proprietary 25-year-old chips...not an easy task.
While we're all praying to the Gospel of Altoona and the successful extra lives their endless rebuilds have bestowed on the AC44C6M's and countless old EMD's, remember that the most ambitous upgrade of a Dash 8 product ever undertaken--the Altoona rebuild of ex-Conrail 8-40CW's into 8.5-40CW's with special NS-designed upgrades--was a flop so gigantic that the rebuild program at Altoona was ended after 4 years and only 10 (scaled-back) completed units, with most units getting pulled back into storage after fewer than 18 months of actual road service (most not even that much), and all stored units now expected to be dispersed after ongoing brand-new orders complete. The systems integration of meshing old Dash 8 circuitry with ambitious new eco-control and computer automation proved to be too big a nightmare. Right now the rest of the 8-40CW's on NS's active roster have no immediate plans for any major rebuild and have just gotten some light emissions updates in the form of physical mods for radiators and a new fuel injection system non-dependent on any circuitry changes.
Dash 8's on other Class I's (cobbled mostly from
The Diesel Shop site). . .
-- CSX had a much larger roster of ex-Conrail Dash 8's to begin with, but has opted to place most of their B40-8's into off-roster storage and start selling them to Class II's & III's (see Pan Am's recent pickups), leaving just a large late-era production run C40-8W's that are just now sitting at 25-year rebuild age awaiting some sort of decision on what comes next.
-- BNSF has started scrapping some of its B40-8's as parts donors to keep the rest of its active fleet running in-situ, instead of calling for rebuilds. Their C40-8W roster is mostly stored.
-- UP still appears to run a large number of C40-8's, C40-8W's, and C41-8W's. Doesn't appear, however, that any of theirs have been refurbished outside of factory spec.
-- CN has a number of C40-8's acquired secondhand from UP, C40-8W's acquired secondhand from BNSF, and C40-8M's + C40-8W's that came over from BC Rail and Illinois Central when they absorbed those roads. Other than some minor de-ratings within GE factory spec while still at UP, doesn't appear anything's been changed on them while with CN.
-- CP, KCS, and Ferromex don't appear to have any rostered Dash 8's (though Ferromex seems to be quite upgrade-happy with its Dash 7's).
The numbers don't lie. A huge sample size exists of this class with the Class I's, and they are
not getting the future-leaning upgrades of their second-gen microprocessor successors or their analog control predecessors. Nothing exists on that platform that's anywhere close to the scope of the AC44C6M, or even the near-endless supply of eco-upgraded GP38-2's that Altoona keeps pumping out from old Geep 38/40/50 bodies. Most of the Class I-rostered Dash 8 makes are running out their 25-year pre-rebuild lifespans with no major overhauls in sight (on roads that are actively doing extremely major eco-'n-otherwise upgrades on other power), or being maintained to in-situ state-of-repair with little to no changes in factory spec and a whole lot of parts donations from other sacrificed units. The Class II's that are starting to pick up Dash 8 dispersals from Class I storage (PAR, NYSW, longstanding users P&W, etc.) are likewise maintaining them to in-situ spec commensurate with their smaller fleet management resources
----------------------------------------
The P42 (or 8-42BP in Dash 8 numbering parlance) has the second-most sophisticated circuitry of any Dash 8 make, trailing only the
hyper-custom P32AC-DM and its whole extra bag of tricks. The extra overhead for managing HEP, optimal passenger performance, and the systems reliability requirements of a passenger loco made those circuit boards significantly more complicated than their earlier freight counterparts. And more sophisticated than the P32-8BWH with its more conservative performance profile. All of the same factors that limit the freight Dash 8's upgrade abilities are in-play here, and then some. So before assuming that the P42's can just be "Altoona'ed" on-a-whim, remember the lesson NS bitterly learned with the 8.5-40CW debacle: it is very damn hard to upgrade around mid-90's hard-wired logic boards with new computer doohickeys and make it all sing coherently. Your options for rebuild are either:
1. In-situ, like most of the light SGR refreshes the freight Dash 8's have gotten, and (so far) the in-house refreshes on the P32-8BWH. Which means solving the problem on how to reproduce those now-unsupported circuit boards on Gennie units whose boards are a hell of a lot more complicated than your average C40-8W road freight that the Class I's aren't exactly rebuilding. CDOT can make that work for them unobtrusively by rebuilding a mere half-dozen P40's and banking on parts availability from AMTK's P40 fleet being at the front of the retirement line once all 63 currently ordered Chargers are on the road (at the very least the 14 stored non-reman units will probably be scrapped as parts donors for the rest of the Gennies, allowing additional acquisition of spare replacement boards). But there were plenty of horror stories recounted on RR.net about what a pound of flesh the P32AC-DM rebuilds took out of Metro North around the tapped-out supply chain on those boards. Amtrak could do this if they only needed a few dozen Gennies as supplementals after the Charger options were drained, because they'd have the whole of 150 or so stored units to cannibalize for parts to keep the others fresh and the hardest-to-produce parts like the circuit boards well-stocked in a warehouse for 10+ years of service. Ditto for the commuter rail operators who pick through the AMTK dispersals...if 1 unit gets cannibalized for every 5 that continue running there's probably enough extra boards to go around to plot a CDOT-like full in-situ overhaul for 20-year extended lifespan. But you'll get barely any emissions improvements, and nothing else in the way of substantial upgrades because it's rehab bottled-up totally within original factory spec.
2. Compatibility layers glommed onto the old boards to add new functions and make bypasses for replaced functions. This is exactly what NS tried...and failed...to do with the 8.5 remans. And it also doesn't solve the supply chain problem with the now-unsupported original boards. Given just how tightly-integrated the Gennies' systems are, this is going to be nigh impossible to do effectively.
3. Reverse-engineer the Gennies' controls as firmware microcode, so all subsequent mods can be done as "software" just like current-gen power. This is not nearly as easy as adding microprocessors to a formerly analog F40PH-2 to make it an emissions-controlled F40PH-3 or other such upgrades. Highly-custom hardware pathways have to be painstakingly re-created in the firmware, without a hint of difference to the systems that have to interact with it. If GE doesn't spend the R&D to do that extremely tough and thoroughly unrewarding-to-self porting job on boards that are still its proprietary intellectual property, there isn't a straightforward path for some third-party to produce Genesis-clone firmware and debug the thing well enough to be production-ready. None of the freight Dash 8's have had this done to make them as forward-upgradeable as the preceding and succeeding generations of GE power, and Metro North didn't have this option available to them when rebuilding its P32's...so before talking up a reverse-engineered P42 computer consider where the numbers
ain't in the rest of Dash 8-land. It's not like you can "boot up" a virtualized Gennie computer in emulation as if it's been cued up like some old version of Windows in Parallels Desktop on your Mac.
4. Outright replace the Gennies' controls with a total, from-ground-up firmware rewrite that can handle the Gennie parts you want to keep (from 7FDL on down), enables the parts you want to swap (true state-of-art eco controls, etc.), but is fundamentally a new system more in line with modern power and shorn of all the limitations of primitive mid-90's microprocessor tech. Note also for the numbers game that this was never considered an option for freight Dash 8's, nor does it appear that any of the Class I's still holding onto stored or non-rebuilt fleets are seriously considering something like this. That should speak volumes about the costs involved designing such a system, since software is now the most intensive and expensive part of the design process for all-new power and is more what makes the loco make its own distinct make than the shopping list of mechanical parts stuffed inside. Does it make sense to basically do R&D for an entirely new loco to create this new computer...but surround it with entirely old parts??? That's pretty much what you have to do to have the easiest path at grabbing one of the lower emission tiers for that 7FDL-16 prime mover...because change anything major like DC-to-AC traction and the EPA kicks you immediately up to a Tier 4 requirement on the rebuild. Might as well go whole-hog with a GEVO and gut + re-kit that monocoque shell to something entirely new and
other if those are now the terms of engagement. This may indeed be an option, especially for commuter rail operators and third-party builders who see opportunity in 150-190 monocoque hulks up for dispersal. But the resulting product won't be a Genesis in anything except donated carbody. For Amtrak to make that price point work for them in-house it basically has to put 150+ P42's through this program to pay down all that up-front R&D with the unit cost discount from very high number of ordered units. That means banking on an assumption that zero or near-zero of the national Charger options get picked up, and that somehow an RFI can be issued establishing equal-or-lower price point before a decision has to be made on the sure-thing price they already have on the Charger options. If you've been on the side of thinking the QSK19R prime mover is too big a risk to prove itself in LD service you may throw more chips in this particular betting pool. Just keep in mind that the Cummins engine is now a real revenue-service product, and every additional day that goes by without major complaints from the mechanical dept. deflates that risk a little bit more and makes the economic argument of "locked-in price" vs. "who knows...we don't even have an RFI???" harder to sustain for this rebuild option. Plausible deniability cedes a little more ground every additional day the Cummins doesn't prove its doubters right.
