July TRAINS item on electrification - the "FL9" solution?

[Anonymous]

I take it also that the idea of an abreviated electrification was never studied, aka instead of keeping the electrification from Avery to Harlowtown, reduce it to Avery-Haugen/Butte-Whitehall/Ringling-Martinsdale respectively.  Again, it comes down to the cost of maintaining catenary where needed (with that 5,400 hp per unit under wire) but eliminating it where diesel mode (at 3,000 hp per unit) would suffice.

The cost of building the catenary was an obstacle, but not the cost of maintaining it.

I remember vaguely one of your posts from a while back when you discussed the decision making involved in the eventually abandonment of the Milwaukee electrification.  You had worked on this first hand if I remember correctly, and the analysis you and your coworkers came up with showed that maintaining the electrification was preferable to abandoning it, yet management chose the latter supposedly due to the economics of standardization.

One question:  What was the analysis of the electric locomotive situation vis-a-vis remaining useful service life of the Joes et al and the cost of buying new electrics to replace the 50 year old boxcabs during the discussion of whether to continue the electrification or not?  Can we assume the Joes still had 20 or 30 years left in them in 1974?  What about the cost of replacing the older electrics (assuming they needed replacement)?

One thing I'm getting at is that a dual mode locomotive would not necessarily have been superfluous to the locomotive market if electrification was ended, since they could be stripped of their electrification equipment and still ran as straight diesel-electrics.  But if Milwaukee bought brand new electrics, they were stuck with them with no resale possibility if electrification was abandoned a few years later.

Another comparitive consideration is the recent plethora of new locomotive designs, aka hybrids, gen-sets, and now this new GE prototype where the power from dynamic braking is stored in battery packs rather than disapated as heat.  Why would a modern day dual mode locomotive be any more of a maintenance hassle than these current offerings?  Again, if the one thing keeping modern electrification from making a return is the up front costs of electrifying whole subdivisions, wouldn't segmented electrification of these subdivions and whole fleets of dual mode locomotives be a less expensive option than wholesale electrification and whole fleets of single mode electric locomotives?  Wouldn't dual mode locomotives have better resale prospects down the road than new electrics?

One of the compelling arguments General Electric made to MILW in 1972 for completing the "Gap" was that the use of a continuously electrified section would increase the normal efficiencies associated with electrification. With the two separated sections, maintenance costs per hp were about 47% of the costs of maintaining an equivalent Diesel-electric horsepower. This was high for a heavy electrification and due to several factors including the cost of operating two separate maintenance facilities, two separate fleets, and relatively shorter runs by the electric locomotives. By making a continuous section, and as long as possible, GE estimated that MILW could get its electric power maintenance costs down to the expected 30% of equivalent Diesel-electric horsepower.

I agree that electrifying the gap made sense as the distance between necessary stops was lengthened with time and locomotives could stay with consists conceivably across the country.  But there's still the issue of standardization.

I have the GE Econometric program used in the study, and as a "for example" input the cost of a dual mode locomotive and compared that to the cost of full Electrification, Harlowton to Tacoma, and compared that to full Dieselization as well, making a variety of assumptions regarding growth rate of traffic, inflation, and utilizing real costs of diesel fuel and electric power, 1974-2004, assuming a 15 year economic service life of the Dual Mode locomotives, and for a regular Diesel-electric, and 30 years for the straight electric,

Why assume only a 15 year economic service life for the dual mode?  Didn't the FL9's last 50 years? 

If we assume 15 years for diesel and 30 years for electric, wouldn't a more logical dual mode assumption be 20 to 25 years since it is operating part time as an electric and part time as a diesel?  The other consideration is that it is operating in diesel mode over less strenuous territory and in electric mode over the more strenuous territory, so it's diesel service life portion should last longer than diesels that went everywhere.

[Tom Curtin]

Some comments on the history and evolution of the FL9

1. It is true that the initial idea by NH management for a dual power locomotive back in the early 50s was to take some load off the railroad's Cos Cob, CT power plant. But by the time the EMD FL9's were actually ordered in 1956 this strategy had changed completely because the RR was under control of a severely misguided management that considered the whole idea of electrification obsolete and was determined to replace all existing diesel and electric passenger power with FL9s.

2. NH was not short of electric locomotives when they ordered the FL9s. In fact they retired a number if fine straight electrics before their time.

3. Both FM and Alco issued to the NH a proposal for a dual power locomotive. However, at the time the strategy was under the control of a manager who was fanatically pro-EMD and wouldn't hear of any other builder's designs. No looking for bears behind trees on that one, those are the facts, pure and simple.

4. The FL9's have had an utterly amazing career for a number of owners. Really amazing when you consider how badly they were neglected in midlife by a couple of bankrupt owners.

5. Before Metro North replaced them with the GE "Gennie" units the DC power option was non-functional on just about all of them, and they were operating on diesel into GCT, despite laws to the contrary. A few are still running on MN today, but not into GCT.

Despite their long service lives, all the research on their history supports that they were originally born due to a totally misguided strategy.   The New Haven blew away tons of capital on this --- capital they didn't have in the late 50s, and no doubt that strategy hastened the company's bankruptcy.

 

[MichaelSol]

You're pretty savy with economics and accounting, but leave the engineering to those who practice it!

Ha!

It is true that I stopped renewing my membership in the Surface Chemistry section of the American Chemical Society in 1984. I was practicing law at that point, and wasnt't going back into the field. However, during my ten years with the USDA, and my reason for having that particular membership, was because one of my odd duties -- didn't consume a lot of day to day work -- at USDA was to be a primary investigator at the Metallurgical Fatigue and Surface Chemistry Laboratory. Tell me all about the engineering and stress on metal of vibration and environment.

My supervisor only had a BA in chem, whereas I was finishing a doctorate in chemistry and had the bachelor Chem E and so she did the administering and I did the chemistry. This involved study of surface erosion dynamics and stress application to various metals and alloys in various environments, determining effects both by periodic examination through hundreds of thousands of stress cycles using a Carl Zeiss Metallurgical microscope as well as measurement to failure through fatigue testing under varying levels of stress.

You are just guessing, whereas I do, in fact, bring a specific background in metallurgical analysis and testing to my conclusion based on specific professional experience in design, testing and evaluation of metal fatigure and surface erosion.

I'm not going to write a dissertation here, nor respond specifically to your condescending remark, but you almost have the key, you are just holding it backwards.

It is indeed a matter of energy.

And you do not, nor will you ever, have the sample necessary to reach the conclusion you feel strongly about, unless somebody dedicates a diesel locomotive to hauling around, starting it up and shutting it down two or three times a day while rolling it a few hundred miles at a time between shutdowns, and doing this for between 8 and 30 years. However, energy will transfer to cylinder walls and bearings. It's slow, but it will happen from normal vibration. And it will cause site specific erosion, particularly if the engine, as many do, stop in a specific configuration each time. And with a dual purpose locomotive, you would have many more hundreds of thousands of such cycles than you would ever have in the normal operation of an occasional tow of a locomotive.

 

[MichaelSol]

Question for Michael Sol:  Were the box-cab electrics on MILW also equipped to run in multiple with diesels in a fashion similar to the Joes? 

Yes, a similar system was developed by H.R. Morgan in 1957 and allowed the Boxcabs to MU with Diesel-electrics.

 

[oltmannd]

Too, there is a vibration problem. The equipment that is not being used doesn't just sit there while the other subsystem is operating. In electric mode, the lubricants in the diesel engine are all being shaken down off the cylinder walls and out of the bearings. And bearings don't like to just sit in one spot and vibrate. Neither do piston rings. In diesel mode, all those important electrical contacts are just sitting there, breathing the nice moist or dirty air for 1,000 miles of their journey.

Non-issues.  RRs regularly tow dead power hundreds of miles at a shot with no adverse effect.  Surface tension of the oil film will hold plenty of oil where it's need for quite a while.  You only have to pre-lube the engine after it's been dead for 2 days or more. 

I have no doubt that there is a difference between a one-time tow, for however many hundreds of miles, and a daily characteristic of operation. Can't see an analogy between a 400 mile tow once in a while, and a regular daily 600 mile run, up to 200,000 miles per year. These just aren't comparable. There is a 50,000% difference in the cycles, assuming one tow cycle per year, compared to the operating cycles.

Even if the tow cycle was once a month, the engine would have to last 4,166 years to experience the equivalent exposure to non-operating vibration and buffeting motion that the engine experiences in a single year of normal operation.

This is an extremely poor sampling comparison and simply cannot be the basis for a valid conclusion.

The diesel engine itself is a much larger source of damaging higher frequency vibration than transmitted through the suspension.  A half a G at 1/2 to 2 Hz doesn't have enough power to hurt anything.

This is refering to an operating locomotive, operating on a trolley while the diesel engine is off.

The switch between electric and diesel on an AC traction dual mode loco would be similar to a rotary DB switch on an EMD or set of contactors on a GE on a current diesel loco.  A few days of non-use in a filtered, pressurized electrical cabinet wouldn't hurt'em even a little bit - particularly compared to what happens to them under normal use when they break an arc.

Ahhh .. "filtered" ... "pressurized." Code words for "filter not changed," "seal broke", "compressor failed." More parts. "A few days." Again, this doesn't recognize the daily operating cycle of day in and day out. I have no doubt it would be better than the "open" cabinets of the GE Boxcabs and Joes, even at 3,400 volt DC, even opening and closing constantly compared to continous operation, and even in extremes of service conditions.

 

OK.  Lets try an analogy.  Using a set of GE contactors to switch between the traction alternator and traction transformer on a dual mode locomotive would be like using locomotive battery knife switch to operate a flashlight.

Trying to damage engine parts letting a shut down diesel travel over the RR is like trying to mix paint by putting the can in the trunk of a Rolls Royce and driving on an interstate.  It's about energy.  You can tow power dead on a 50% duty cycle, a day off and a day on, year in and year out an not do a bit of damage.  The lubrication at start up will be just fine. 

You're pretty savy with economics and accounting, but leave the engineering to those who practice it!

[CSSHEGEWISCH]

FM must have a real taste for irony since the Speed Merchant (P12-42) was an even more specialized application than the FL9.  Only four of them were built (2 for NH, 2 for B&M) and only those on NH had third-rail shoes.  They were not suitable for any purpose other than pulling Talgo or similar trains.

Similarly, the Baldwin RP210H used a torque-converter drive to presumably eliminate many of the complexities of a standard electric drive.  The addition of third-rail shoes and other control equipment to the NH power negated this theoretical advantage.

Question for Michael Sol:  Were the box-cab electrics on MILW also equipped to run in multiple with diesels in a fashion similar to the Joes? 

[MichaelSol]

I take it also that the idea of an abreviated electrification was never studied, aka instead of keeping the electrification from Avery to Harlowtown, reduce it to Avery-Haugen/Butte-Whitehall/Ringling-Martinsdale respectively.  Again, it comes down to the cost of maintaining catenary where needed (with that 5,400 hp per unit under wire) but eliminating it where diesel mode (at 3,000 hp per unit) would suffice.

The cost of building the catenary was an obstacle, but not the cost of maintaining it.

One of the compelling arguments General Electric made to MILW in 1972 for completing the "Gap" was that the use of a continuously electrified section would increase the normal efficiencies associated with electrification. With the two separated sections, maintenance costs per hp were about 47% of the costs of maintaining an equivalent Diesel-electric horsepower. This was high for a heavy electrification and due to several factors including the cost of operating two separate maintenance facilities, two separate fleets, and relatively shorter runs by the electric locomotives. By making a continuous section, and as long as possible, GE estimated that MILW could get its electric power maintenance costs down to the expected 30% of equivalent Diesel-electric horsepower.

I have the GE Econometric program used in the study, and as a "for example" input the cost of a dual mode locomotive and compared that to the cost of full Electrification, Harlowton to Tacoma, and compared that to full Dieselization as well, making a variety of assumptions regarding growth rate of traffic, inflation, and utilizing real costs of diesel fuel and electric power, 1974-2004, assuming a 15 year economic service life of the Dual Mode locomotives, and for a regular Diesel-electric, and 30 years for the straight electric, as well as a 9% annual interest charge or lease service charge. This assumed a 2.7% annual traffic growth over 30 years, whereas the actual traffic growth in the electrified territory, 1970-1977, averaged 8% per year, and had averaged 5% per year, 1960-1970. The significance of this point is that, again, an unjustified conservatism in the use of estimates grossly underestimated the impact of the decision, and in this case, the higher growth percentages strongly leveraged the outcomes in favor of the Electrification, but those numbers were not generated by the studies, and the executives reviewing the studies did not have a realistic range of estimates presented to them.

In addition, the numbers generated below were not based on the numbers used by the studies, but rather the actual power and fuel costs that existed over the thirty years since the studies were made. The studies, in hindsight, grossly underestimated the fuel cost changes, and grossly overestimated the price changes in electric power. Indeed, the studies used an average increase of 1% per year in electric power costs, even while acknowledging that, historically, costs had continually declined.

In these comparisons, the lease or interest charges on the locomotives is included in both the capital cost calculation and the operating cost, as the program was looking at "annual cost" of operation being the total cash flow, and full cost of financing charges (compared to NPV) rather than "operating" expense from a strictly accounting standpoint.

Capital cost, 30 years, NPV

Dual Mode SD40-2    $56,776,072

Full Electrification     $23,781,752

Full Dieselization       $39,417,480

Total 30 year Operating Costs (fuel, power, interest/lease charges, maintenance, inventory, inspection, lubricants)

Dual Mode SD40-2    $1,137,396,413

Full Electrification        $364,418,585

Full Dieselization        $1,048,739,950

Anything that would have diminished the role of the straight electric would have just cost money. And it's just hard to extract any operating savings from the Dual Mode system because of the higher capital cost and financing charges not only cost more, but welded that system to the rapidly escalating diesel fuel charges of the next 30 years, compared to electric power costs which actually declined from $.049/kwh in 1974 to $.048/kwh in 2004 (industrial rate), while diesel fuel went from $.096/gal. to $1.64, a 1700% increase.

Now, a qualification. These numbers are generated for an existing electrification, contemplating a 212 mile extension or supplementation. It is system specific and includes a variety of cost replacement factors for existing trolley poles and rectifier supplementation necessary for full electrification, signal alterations necessary for full-dieselization, etc, and includes here, for instance, a substantial credit to the capital costs of full dieselization/DM of the salvage value of the copper in the existing system, as well as a similar credit to the cost of full-electrification of the depreciated cost of existing Diesel-electrics that would be released for service elsewhere. It is unique for the specific system at the specific point in time for its specific needs. These are not the capital costs of electrifying the resulting 880 mile system from scratch and can't be seen as "showing" electrification as cheaper than alternative systems for any kind of a generalized system from an investment standpoint, as that would require a different calculation.

 

[MichaelSol]

Well, that's a good example, albeit I can see a bit of an aberation regarding the westbound Joe into Avery being the same Joe for the eastbound out of Avery.  If that's the way it was done, then so be it, but was it normal SOP for the inbound Joe at Avery on #261 to also be the outbound Joe for #262? 

Oh, I doubt it was mandatory. On March 29, 1973, notes show that E78 came in on #261TC27 at 12:01 p.m., was turned, inspected, and left on #264, a 5500 ton train, at 4 p.m.. Happened to be that #264 was the next train out, rather than #262. E71 was placed on Advance 262S29 as it was leaving Avery about 40 minutes before #261 arrived.

However, in the example of the four hotshot trains each way, the fleet numbers of 32 SD40-2s and 4 Joes were what made it work. The individual SD unit ID numbers or Joe ID numbers weren't necessarily relevant from the standpoint of making an economic calculation, as specific locomotives were also cycling through other trains as well.

 

[Anonymous]

On the FL9 versus the Fairbanks Morse engine issue - I remember reading somewhere (probably TRAINS) that EMD threatened New Haven fairly overtly that if they chose the Fairbanks Morse engine, a lot of the GM cars going by New Haven would be switched (to trucks I guess).

Again, the FL9 eventually proved itself to be a hell of a locomotive, but the story, if true, speaks to GM's (and later GE's) ability to sway buyers their way.

ALCO, Baldwin, Fairbanks Morse, and others, weren't also huge railroad clients, like GM and GE were.  (Fact that GM made a superior product, also helped, don't get me wrong - but it's an interesting aside.)

Well, that brings me to another supposition.  Since the Milwaukee was an unabashed FM customer, at least for it's Olympian Hiawatha, I wonder if they ever considered producing a version of the dual mode P-12-42 aka "Speed Merchant" that could run under Milwaukee's 3kv catenary, rather than using the Erie-builds the whole way from Chicago to Seattle?  Did the OH ever use Joes as helpers for the FM-powered version, or was the three unit FM consist sufficient across the system?

Would that same cost differential apply regarding the Erie-builds vs the Speed Merchants?

[Anonymous]

The idea of a dual-source locomotive makes little sense. EMD pitched a specific design to the Milwaukee Road in 1972, converting SD40 locomotives to alternative 3kV DC by the addition of a pantograph and related control equipment. Naturally, it cost more than either the straight Diesel or the straight Electric. This meant, when being used in the Diesel mode, it cost more per horsepower than a comparable Diesel-electric. When used in the electric mode, it cost more per horsepower than a comparable straight Electric. On 600-800 miles runs, there just wasn't a cost savings compared to simply switching motive power.

In this analysis, you're continuing to maintain two separate fleets.  But what if the dual mode locomotive allowed for a 1 for 2 replacement, aka 1 dual mode locomotive replaces 1 straight electric and 1 diesel?

Well, a locomotive can only be in one place at one time.

Maybe an example.

Milwaukee Train # 261 generally arrived in Harlowton at 9:45 p.m. powered by 4 SD40 locomotives, 12,000 h.p. At Harlow, a 6,000 h.p. Little Joe electric was put on. The SD40 had a 1,020 ton rating from Piedmont to Donald, the ruling grade on the run, at 18 mph. The Joe was rated at 1,440 tons at 25 mph. on that grade. The train on April 28, 1972 was limited to 50 cars, 3500 tons. The Joe was taken off at Avery and the train ran on with the Diesel engines to Tacoma arriving at 4 a.m. 31 hours later. At Avery, the Joe could be turned and run eastbound on 262 in the afternoon. Basically, it was that Joe running over the three mountain ranges on the Rocky Mountain Division that gave both 261 and 262 their fast times. Train #262 likewise left Tacoma with a four unit SD40.

So, the "train cycle" for #261 and #262 involved two sets of four SD40s, eight total, and a Little Joe Electric that swung between the two trains for the Rocky Mountain Division run. The cost of the train cycle was as follows:

Little Joe equivalent: $540,000

SD40s @$270,000 = $2,160,000.  Total cost for the equipment cycle = $2,700,000.

The dual-mode SD40s were estimated at 140% of the cost of the Diesel-electric version of the SD40, although another estimate was 180%.  Using the 140% estimate, the bad news: in diesel mode they put out 3,000 h.p.. The good news, in electric mode they generated 5,400 hp.. So, #261 would still need four of them to haul the train in Diesel-electric mode, eight for the total cycle. Cost $3,780,000. But, with four of them, at 5,400 hp in the electric mode, the train has 21,600 hp. compared to the 18,000 hp in the combined system. The extra horsepower doesn't really do that much good, but it costs $1 million more to have it there, because you still need the four locomotives to power the train where there is no trolley, but it only cost $540,000 to have it there where it was needed in the form of a Little Joe.

So the cost of the combined system, to maintain the #261/262 cycle was $2,700,000, whereas the cost of of the dual mode SD40 system to achieve the same result was $3,780,000.

A straight electric system would also cost $2,700,000, identical to the combined system, to obtain the necessary horsepower.

Adjusting for availability, the dual mode locomotive gets dim. It would cost as follows to purchase the motive power equipment to operate #261/262 on the Harlowton/Tacoma cycle:

Straight electric: $2,934,783.

Combined system (Milwaukee Electrification): $3,158,385.

Straight Diesel-electric (10 units, Milwaukee Dieselization): $3,214,286. It was a little higher than this as SD45s were thrown into the mix.

Dual mode locomotive system: $4,973,684.

Adjusting for economic service life and financing charges, the systems diverge considerably.

The Power Manual in effect April, 1972 shows 4 sections of #261 operating between Harlowton and Tacoma, and 4 sections of 262.

The total motive power costs are as follows:

Electric: $11,739,132.

Combined System: $12,633,540.

Diesel-electric: $12,857,144.

Dual-mode: $19,984,736.

Well, that's a good example, albeit I can see a bit of an aberation regarding the westbound Joe into Avery being the same Joe for the eastbound out of Avery.  If that's the way it was done, then so be it, but was it normal SOP for the inbound Joe at Avery on #261 to also be the outbound Joe for #262? 

I take it also that the idea of an abreviated electrifiction was never studied, aka instead of keeping the electrification from Avery to Harlowtown, reduce it to Avery-Haugen/Butte-Whitehall/Ringling-Martinsdale respectively.  Again, it comes down to the cost of maintaining catenary where needed (with that 5,400 hp per unit under wire) but eliminating it where diesel mode (at 3,000 hp per unit) would suffice.

That being said, would the #261/262 have needed four diesels on the flatter stretches, or could it have made it with three, knowing that when grades were approached and catenary became available, that 9,000 combined hp under diesel mode would become 16,200 hp under wire?

To bring this to a contemporary example, take the MRL between Helena and Mullan Tunnel.  MRL has been running a helper district here to tackle the 2.2% westbound grade, right?  Also, operation of hard working diesels inside the tunnel is a constant problem.  I think this is an example of where a dual mode locomotive would work out best financially, since it would replace manned helpers.  Just electrify from Helena to the west portal, no need to cut helpers in and out.  Maybe do the same on Bozeman - electrify the eastbound and westbound grades, but leave the flatter sections as is.  That would allow MRL to run straight from Laurel to Spokane without any engine changes, helpers in/helpers out, or having to overpower the consists on the flats to ensure adaquate hp for the grades.  Unless and until someone invents the fully automatic unmanned DPU.......

Although I suspect Michael would suggest electrifying the MRL from Garrison to Livingston!

[DMUinCT]

Just one more point. The FL9 never took power from the catenary, only the third rail.

It ran on diesel power in Catenary Territory from New Haven CT to Mt Vernon NY and then dropped the 3rd rail pickup shoe wher the Catenary ended. The shoe dropped and retracted under power, if the engineer remembered.

[DMUinCT]

The FL9s, built for The New Haven, went over to the Metronorth Commuter Railroad, CT Dept. of Transportation, and Amtrak. They remained in service, working passenger trains, to Grand Central Station from Albany, Waterbury, New Haven, and Danbury for 50 years.

Other Main Line locomotives were traided off to Local and Regional railroads to continue to work for years on lighter duty.

[GP40]

Futuremodal,

I believe that max speed of the FL9 in the Wikipedia entry is wrong and kind of low.  The FL9's had a 59:18 gear ratio and with their old D55 or D67 traction motors that would have given them a max speed of around 83 MPH not just 70 MPH.

[MichaelSol]

Too, there is a vibration problem. The equipment that is not being used doesn't just sit there while the other subsystem is operating. In electric mode, the lubricants in the diesel engine are all being shaken down off the cylinder walls and out of the bearings. And bearings don't like to just sit in one spot and vibrate. Neither do piston rings. In diesel mode, all those important electrical contacts are just sitting there, breathing the nice moist or dirty air for 1,000 miles of their journey.

Non-issues.  RRs regularly tow dead power hundreds of miles at a shot with no adverse effect.  Surface tension of the oil film will hold plenty of oil where it's need for quite a while.  You only have to pre-lube the engine after it's been dead for 2 days or more. 

I have no doubt that there is a difference between a one-time tow, for however many hundreds of miles, and a daily characteristic of operation. Can't see an analogy between a 400 mile tow once in a while, and a regular daily 600 mile run, up to 200,000 miles per year. These just aren't comparable. There is a 50,000% difference in the cycles, assuming one tow cycle per year, compared to the operating cycles.

Even if the tow cycle was once a month, the engine would have to last 4,166 years to experience the equivalent exposure to non-operating vibration and buffeting motion that the engine experiences in a single year of normal operation.

This is an extremely poor sampling comparison and simply cannot be the basis for a valid conclusion.

The diesel engine itself is a much larger source of damaging higher frequency vibration than transmitted through the suspension.  A half a G at 1/2 to 2 Hz doesn't have enough power to hurt anything.

This is refering to an operating locomotive, operating on a trolley while the diesel engine is off.

The switch between electric and diesel on an AC traction dual mode loco would be similar to a rotary DB switch on an EMD or set of contactors on a GE on a current diesel loco.  A few days of non-use in a filtered, pressurized electrical cabinet wouldn't hurt'em even a little bit - particularly compared to what happens to them under normal use when they break an arc.

Ahhh .. "filtered" ... "pressurized." Code words for "filter not changed," "seal broke", "compressor failed." More parts. "A few days." Again, this doesn't recognize the daily operating cycle of day in and day out. I have no doubt it would be better than the "open" cabinets of the GE Boxcabs and Joes, even at 3,400 volt DC, even opening and closing constantly compared to continous operation, and even in extremes of service conditions.

 

[oltmannd]

I believe the combined availability number would be correct for both subsystems (Diesel-electric and electric) to be simultaneously operable (system availability).  This would be the number to use if the combined unit had to use both the Diesel-electric subsystem and the electric subsystem at the same time.  If the combined unit used only the electric subsystem under catenary and only the Diesel subsystem while not under catenary, then one has to take into account the fraction of the trip under catenary to get a detailed estimate.  Depending on what that fraction was, the availability of the combined unit for the trip would be somewhere between 84% and 92%.

If Diesel operation was permitted under catenary, although not preferred, then the Diesel subsystem would have been a backup to the electric subsystem.  In that case the likelihood of the engine being inoperable while under catenary was .08 x .16 =  .0288.  So the availability of the combined unit under catenary was 97%.  The electric subsystem is not relevant when not under catenary, so the availability when not under catenary was just the availability of the Diesel subsystem, or 84%.  In this case the availability for the trip would have been somewhere between 84% and 97%. 

Too, there is a vibration problem. The equipment that is not being used doesn't just sit there while the other subsystem is operating. In electric mode, the lubricants in the diesel engine are all being shaken down off the cylinder walls and out of the bearings. And bearings don't like to just sit in one spot and vibrate. Neither do piston rings. In diesel mode, all those important electrical contacts are just sitting there, breathing the nice moist or dirty air for 1,000 miles of their journey.

Non-issues.  RRs regularly tow dead power hundreds of miles at a shot with no adverse effect.  Surface tension of the oil film will hold plenty of oil where it's need for quite a while.  You only have to pre-lube the engine after it's been dead for 2 days or more. 

The diesel engine itself is a much larger source of damaging higher frequency vibration than transmitted through the suspension.  A half a G at 1/2 to 2 Hz doesn't have enough power to hurt anything.

The switch between electric and diesel on an AC traction dual mode loco would be similar to a rotary DB switch on an EMD or set of contactors on a GE on a current diesel loco.  A few days of non-use in a filtered, pressurized electrical cabinet wouldn't hurt'em even a little bit - particularly compared to what happens to them under normal use when they break an arc.

[Erie Lackawanna]

On the FL9 versus the Fairbanks Morse engine issue - I remember reading somewhere (probably TRAINS) that EMD threatened New Haven fairly overtly that if they chose the Fairbanks Morse engine, a lot of the GM cars going by New Haven would be switched (to trucks I guess).

Again, the FL9 eventually proved itself to be a hell of a locomotive, but the story, if true, speaks to GM's (and later GE's) ability to sway buyers their way.

ALCO, Baldwin, Fairbanks Morse, and others, weren't also huge railroad clients, like GM and GE were.  (Fact that GM made a superior product, also helped, don't get me wrong - but it's an interesting aside.)

[MichaelSol]

I believe the combined availability number would be correct for both subsystems (Diesel-electric and electric) to be simultaneously operable (system availability).  This would be the number to use if the combined unit had to use both the Diesel-electric subsystem and the electric subsystem at the same time.  If the combined unit used only the electric subsystem under catenary and only the Diesel subsystem while not under catenary, then one has to take into account the fraction of the trip under catenary to get a detailed estimate.  Depending on what that fraction was, the availability of the combined unit for the trip would be somewhere between 84% and 92%.

If Diesel operation was permitted under catenary, although not preferred, then the Diesel subsystem would have been a backup to the electric subsystem.  In that case the likelihood of the engine being inoperable while under catenary was .08 x .16 =  .0288.  So the availability of the combined unit under catenary was 97%.  The electric subsystem is not relevant when not under catenary, so the availability when not under catenary was just the availability of the Diesel subsystem, or 84%.  In this case the availability for the trip would have been somewhere between 84% and 97%. 

Part of the availability statistic measures frame, carbody, truck, brake service which is common to both systems. And this should mitigate to some extent the statistical 76% availability measure. However, the unit cannot exceed the statistical availability of the Diesel-electric locomotive. And this is part of the problem of a dual-mode system -- it loses the advantages that the straight electric provides, and in that case, what's the point? That unit needs to be refueled and lubricated. An Electric locomotive that needs to be refueled and the engine lubricated, even for part of its journey, loses key advantages of Electrification.

"Availability" is not a measure limited to in-service breakdowns, but the time necessary for refueling, inspection, lubrication, running repairs, general maintenance, breakdown repair and overhauls. The amount of actual time "saved" or "lost" by using the diesel engine under the catenary in the event of an electrical system failure or losing a pantograph is miniscule compared to the general circumstances that contribute to the overall "availability" statistics.

Too, there is a vibration problem. The equipment that is not being used doesn't just sit there while the other subsystem is operating. In electric mode, the lubricants in the diesel engine are all being shaken down off the cylinder walls and out of the bearings. And bearings don't like to just sit in one spot and vibrate. Neither do piston rings. In diesel mode, all those important electrical contacts are just sitting there, breathing the nice moist or dirty air for 1,000 miles of their journey.

I suspect, and it is only a suspicion unfueled by any studies I have seen, but based to some extent on hydroelectric generator observations, that the maintenance costs will be higher, and the availability lower, for both the diesel-electric subsystem and the straight electric subsystem -- because neither one is being used full-time -- and that the statistical availability of 76% might have proven optimistic in actual service.

 

[beaulieu]

One point not caught by the people who suggest that a problem with the Diesel-Electric portion would not prevent using the locomotive as a straight electric, is things like changing out a power assembly while the locomotive is going down the tracks, I don't think so. You wouldn't be able to repair the locomotive while its is being used in the other mode.

[CSSHEGEWISCH]

7. After de-bugging, they ran well, and, with multiple rebuilds lasted 50 years in Main Line Service. Know of any other diesel has done that?  Today, the are still available as reserve power when needed, meanwhile being used by local tourist lines.

Any number of GP7's and GP9's have accomplished the same feat, consider IC's Paducah rebuilds.  The SW14 rebuilds from Paducah were originally NW2's, SW7's and SW9's, and many have resold for service to short lines and industrial users.

[DMUinCT]

The FL9 was a "one of a kind" group of locomotives built to The New Haven Railrod spec.

 It's the 1950s

1.  Hourly train service Boston to New York required a switch of locomotives from Diesel to Electric at New Haven.  A loss of 20 minutes on a 4 hour schudule. The DL109s have long gone to Commuter Service.  The main line is ruled by PA1s amd FMs, always operating in pairs.

2.  The Power Station for the New Haven's over head wires (11,000 volt 25 cycle AC) was at Cos Cob, CT. It was built in 1903. It was long over due for replacement.  Added power was being bought from Conn Edison (NY) and United Illuminating (Bridgeport).

3. All Electric Locomotives dated to the 1930s or before.    EXCEPT, the brand new, 4000 HP EP5 Rectifier Electrics from GE designed to move passenger trains from New Haven to New York at 80 mph.

4. The president of the New Haven has left, the new President is the company "bean counter".  Is there any option to save replacing Cos Cob?  Enter EMD.  They have never bought from EMD!

5.  EMD can make an F9 that could run from the New York Central 660 DC third rail into Grand Central Station (Terminal).  

6. Big Problem, The axel loading is much too high for the 125th Street Viaduct that leads to the Park Ave. Tunnel.  EMD will make a new model, a 5 axel diesel!!  Four wheel front truck, six wheel rear truck and add HEP. We have the FL9 (L for long). Running in pairs, back to back, 4000 HP at 80 mph.  The same as an EP5 electric.

7. After de-bugging, they ran well, and, with multiple rebuilds lasted 50 years in Main Line Service. Know of any other diesel has done that?  Today, the are still available as reserve power when needed, meanwhile being used by local tourist lines.

8. Oh yes, another big problem, The New Haven had "long trem" contracts with the two local power companies. The still had to pay for the unused electric power to the overhead wires. To make some use of this power, The New Haven bought the "second hand" Virginian E33s freight locomotives from the N&W.

9.  The switch from 25 cycle AC to commercial 60 cycle (Hertz) power from local power companies was the death of the northheast electric locomotives (including the GG1), (except rectifier electrics that could be modified).  Today, the Northeast is ruled by "The Acela", the HHP8, and the older AEM7.

 

 

[CSSHEGEWISCH]

While I'm rarely in agreement with Michael Sol, I'd like to thank him for quantifying quite thoroughly what many of us have been trying to say, that dual-mode isn't all that it's cracked up to be.  In the case of the FL9 and the P32's, dual-mode could be considered an expensive addition to cover a specialized situation.  Note that a prior post indicates that Amtrak's operating rules state that the P32's operate in diesel mode when they're outside the tunnels and station.  I would think that Metro North has similar rules.  Others have pointed out the mechanical complexity (higher maintenance costs) involved in a dual-mode design.

Dual-mode is like a lot of other ideas:  it looks good on paper but falls flat in the real world.

[Kevin C. Smith]

Maybe an example.

Milwaukee Train # 261 generally arrived in Harlowton at 9:45 p.m. powered by 4 SD40 locomotives, 12,000 h.p. At Harlow, a 6,000 h.p. Little Joe electric was put on. The SD40 had a 1,020 ton rating from Piedmont to Donald, the ruling grade on the run, at 18 mph. The Joe was rated at 1,440 tons at 25 mph. on that grade. The train on April 28, 1972 was limited to 50 cars, 3500 tons. The Joe was taken off at Avery and the train ran on with the Diesel engines to Tacoma arriving at 4 a.m. 31 hours later. At Avery, the Joe could be turned and run eastbound on 262 in the afternoon. Basically, it was that Joe running over the three mountain ranges on the Rocky Mountain Division that gave both 261 and 262 their fast times. Train #262 likewise left Tacoma with a four unit SD40.

So, the "train cycle" for #261 and #262 involved two sets of four SD40s, eight total, and a Little Joe Electric that swung between the two trains for the Rocky Mountain Division run. The cost of the train cycle was as follows:

Little Joe equivalent: $540,000

SD40s @$270,000 = $2,160,000.  Total cost for the equipment cycle = $2,700,000.

The dual-mode SD40s were estimated at 140% of the cost of the Diesel-electric version of the SD40, although another estimate was 180%.  Using the 140% estimate, the bad news: in diesel mode they put out 3,000 h.p.. The good news, in electric mode they generated 5,400 hp.. So, #261 would still need four of them to haul the train in Diesel-electric mode, eight for the total cycle. Cost $3,780,000. But, with four of them, at 5,400 hp in the electric mode, the train has 21,600 hp. compared to the 18,000 hp in the combined system. The extra horsepower doesn't really do that much good, but it costs $1 million more to have it there, because you still need the four locomotives to power the train where there is no trolley, but it only cost $540,000 to have it there where it was needed in the form of a Little Joe.

So the cost of the combined system, to maintain the #261/262 cycle was $2,700,000, whereas the cost of of the dual mode SD40 system to achieve the same result was $3,780,000.

A straight electric system would also cost $2,700,000, identical to the combined system, to obtain the necessary horsepower.

Adjusting for availability, the dual mode locomotive gets dim. It would cost as follows to purchase the motive power equipment to operate #261/262 on the Harlowton/Tacoma cycle:

Straight electric: $2,934,783.

Combined system (Milwaukee Electrification): $3,158,385.

Straight Diesel-electric (10 units, Milwaukee Dieselization): $3,214,286. It was a little higher than this as SD45s were thrown into the mix.

Dual mode locomotive system: $4,973,684.

Adjusting for economic service life and financing charges, the systems diverge considerably.

Thank you for the example-I was seeing what you were saying "through a glass, dimly" but the actual numbers make it much more comprehensible to me.

This is facinating stuff-thanks for the input, everyone.

[cordon]

I believe the combined availability number would be correct for both subsystems (Diesel-electric and electric) to be simultaneously operable (system availability).  This would be the number to use if the combined unit had to use both the Diesel-electric subsystem and the electric subsystem at the same time.  If the combined unit used only the electric subsystem under catenary and only the Diesel subsystem while not under catenary, then one has to take into account the fraction of the trip under catenary to get a detailed estimate.  Depending on what that fraction was, the availability of the combined unit for the trip would be somewhere between 84% and 92%.

If Diesel operation was permitted under catenary, although not preferred, then the Diesel subsystem would have been a backup to the electric subsystem.  In that case the likelihood of the engine being inoperable while under catenary was .08 x .16 =  .0288.  So the availability of the combined unit under catenary was 97%.  The electric subsystem is not relevant when not under catenary, so the availability when not under catenary was just the availability of the Diesel subsystem, or 84%.  In this case the availability for the trip would have been somewhere between 84% and 97%. 

 

 

[MichaelSol]

Too, combining a subsystem with a 92% availbility (the Electric) with a subsystem with 84% availability (the Diesel-electric), the resulting availability of the dual-mode machine was 76%, requiring more such machines to haul the targeted tonnage.

Where did these numbers come from?  Were they drawn from spec analysis, or did EMD actually provide a dual mode prototype(s) for testing?

Again, in the modern context, is it necessarily true that a dual mode loco would have less availability than either straight electrics or diesels?

The availability of the straight electric was determined by GE from actual experience on the Milwaukee with the GE 750 class, with 25 years of operation -- an extraordinarily good availability for an old locomotive. EMD supplied the availability figures to GE from general fleet experience of the SD40/SD40-2. The resulting availability is a statistical measure. The cumulative probability of independent events occuring is additive of the probability of each independent event occuring.

 

[MichaelSol]

The idea of a dual-source locomotive makes little sense. EMD pitched a specific design to the Milwaukee Road in 1972, converting SD40 locomotives to alternative 3kV DC by the addition of a pantograph and related control equipment. Naturally, it cost more than either the straight Diesel or the straight Electric. This meant, when being used in the Diesel mode, it cost more per horsepower than a comparable Diesel-electric. When used in the electric mode, it cost more per horsepower than a comparable straight Electric. On 600-800 miles runs, there just wasn't a cost savings compared to simply switching motive power.

In this analysis, you're continuing to maintain two separate fleets.  But what if the dual mode locomotive allowed for a 1 for 2 replacement, aka 1 dual mode locomotive replaces 1 straight electric and 1 diesel?

Well, a locomotive can only be in one place at one time.

Maybe an example.

Milwaukee Train # 261 generally arrived in Harlowton at 9:45 p.m. powered by 4 SD40 locomotives, 12,000 h.p. At Harlow, a 6,000 h.p. Little Joe electric was put on. The SD40 had a 1,020 ton rating from Piedmont to Donald, the ruling grade on the run, at 18 mph. The Joe was rated at 1,440 tons at 25 mph. on that grade. The train on April 28, 1972 was limited to 50 cars, 3500 tons. The Joe was taken off at Avery and the train ran on with the Diesel engines to Tacoma arriving at 4 a.m. 31 hours later. At Avery, the Joe could be turned and run eastbound on 262 in the afternoon. Basically, it was that Joe running over the three mountain ranges on the Rocky Mountain Division that gave both 261 and 262 their fast times. Train #262 likewise left Tacoma with a four unit SD40.

So, the "train cycle" for #261 and #262 involved two sets of four SD40s, eight total, and a Little Joe Electric that swung between the two trains for the Rocky Mountain Division run. The cost of the train cycle was as follows:

Little Joe equivalent: $540,000

SD40s @$270,000 = $2,160,000.  Total cost for the equipment cycle = $2,700,000.

The dual-mode SD40s were estimated at 140% of the cost of the Diesel-electric version of the SD40, although another estimate was 180%.  Using the 140% estimate, the bad news: in diesel mode they put out 3,000 h.p.. The good news, in electric mode they generated 5,400 hp.. So, #261 would still need four of them to haul the train in Diesel-electric mode, eight for the total cycle. Cost $3,780,000. But, with four of them, at 5,400 hp in the electric mode, the train has 21,600 hp. compared to the 18,000 hp in the combined system. The extra horsepower doesn't really do that much good, but it costs $1 million more to have it there, because you still need the four locomotives to power the train where there is no trolley, but it only cost $540,000 to have it there where it was needed in the form of a Little Joe.

So the cost of the combined system, to maintain the #261/262 cycle was $2,700,000, whereas the cost of of the dual mode SD40 system to achieve the same result was $3,780,000.

A straight electric system would also cost $2,700,000, identical to the combined system, to obtain the necessary horsepower.

Adjusting for availability, the dual mode locomotive gets dim. It would cost as follows to purchase the motive power equipment to operate #261/262 on the Harlowton/Tacoma cycle:

Straight electric: $2,934,783.

Combined system (Milwaukee Electrification): $3,158,385.

Straight Diesel-electric (10 units, Milwaukee Dieselization): $3,214,286. It was a little higher than this as SD45s were thrown into the mix.

Dual mode locomotive system: $4,973,684.

Adjusting for economic service life and financing charges, the systems diverge considerably.

The Power Manual in effect April, 1972 shows 4 sections of #261 operating between Harlowton and Tacoma, and 4 sections of 262.

The total motive power costs are as follows:

Electric: $11,739,132.

Combined System: $12,633,540.

Diesel-electric: $12,857,144.

Dual-mode: $19,984,736.

 

 

 

 

[Anonymous]

It's not often that I have a disagreement with Michael Sol, and granted I'm in over my head on this subject, but here goes.....

The idea of a dual-source locomotive makes little sense. EMD pitched a specific design to the Milwaukee Road in 1972, converting SD40 locomotives to alternative 3kV DC by the addition of a pantograph and related control equipment. Naturally, it cost more than either the straight Diesel or the straight Electric. This meant, when being used in the Diesel mode, it cost more per horsepower than a comparable Diesel-electric. When used in the electric mode, it cost more per horsepower than a comparable straight Electric. On 600-800 miles runs, there just wasn't a cost savings compared to simply switching motive power.

In this analysis, you're continuing to maintain two separate fleets.  But what if the dual mode locomotive allowed for a 1 for 2 replacement, aka 1 dual mode locomotive replaces 1 straight electric and 1 diesel?

BTW - what was the proposed cost of the dual mode locomotive, at least relative to the costs of the electrics and diesels?  In other words, was the cost of 1 dual mode less than the cost of 1 electric and 1 diesel?  Same as?  More than?

In the context of Milwaukee's decision to end electrification due to the expense of catenary rebuild, wouldn't the dual mode have allowed for some limited sections of catenary to be removed (aka Deer Lodge to Haugen), with the subsequent useful parts redistributed to the remaining sections of catenary?

Too, combining a subsystem with a 92% availbility (the Electric) with a subsystem with 84% availability (the Diesel-electric), the resulting availability of the dual-mode machine was 76%, requiring more such machines to haul the targeted tonnage.

Where did these numbers come from?  Were they drawn from spec analysis, or did EMD actually provide a dual mode prototype(s) for testing?

Again, in the modern context, is it necessarily true that a dual mode loco would have less availability than either straight electrics or diesels?

Too, the concept undercut a key advantage of a straight electric locomotive -- the one-third cost of maintenance and the 30-40 year economic service life. Each dual mode locomotive gave up all of the mechanical and service life advantages of the straight electric -- key reasons for electrifying in the first place.

The FL9 had a useful service life running 50 years, didn't it?  What were some of the differences between the FL9 and the dual mode SD's that would account for the SD's having a lesser service life?

Finally, every time one was used in its diesel-electric mode, crucial electric horsepower was unavailable under the wire -- and that was what the fleet investment was for in the first place, and also the limiting factor. It was a poor use of electric horsepower if the whole idea was to get the maximum utilization of system horsepower out of the high cost overhead instead of the high cost Diesel-electric.

I'm a bit confused on this statement - Did the Milwaukee try and run these locomotives in diesel mode while under active catenary?

Again, the railroad had to look at a significantly larger and more expensive overall fleet to meet its needs, at a significantly higher cost per unit, in order to ensure maximization of the use of the catenary, if there was any likelihood at all that a portion of the fleet would be out "somewhere" burning up diesel fuel. And the numbers worked the wrong way there as well: for every diesel horsepower being used in the diesel-electric mode, the company gave up 2 electric horsepower. It made no sense to ever do that.

They cost more, they were in the shop more, and the railroad would have needed more of them.

It truly was a lose-lose proposition.

It seems to me the Milwaukee just had a crappy example of the dual mode concept compared to NH's experience with the FL9.  Granted, if the Milwaukee had no intention of reducing the number of electrified sections to preserve and extend the useful service life of more critical sections of catenary, then the dual mode locomotive concept probably didn't make a whole lot of sense.  It seems Milwaukee's attitude toward overhead wires was to either extend the catenary through the gap between Avery and Othello and thus have a premium electrified railroad from Harlowtown to Puget Sound, or get rid of it altogether.  Would this be a correct assumption?

[Anonymous]

At the risk of being tarred and feathered, I am going to suggest that the FL9 was overrated and undermaintained.  It was proposed as a way of replacing both aging diesels (DL-109's) and aging electrics in one fell swoop.  NH was quite short of operating electrics at the time the FL9's were ordered and the original plan was to operate them in the electric zone in peak periods only to take some of the load off the power plant.  Remember, they were not equipped to run off of the AC catenary.  The dual-power provision was only to allow them to run into GCT, and as the years progressed and proper maintenance lagged, some of them were running into GCT on diesel power.

The FL9 was offered to both PRR and NYC, but neither road purchased them.  Both roads continued to change power at the end of the electric zones.

Currently, both Amtrak and Metro North operate dual-power locomotives but it would be useful to know how much time they actually operate off the third rail.  I would suspect that they cut over to conventional diesel-electric operation as soon as they reach open air, even within the electric zone.  Additionally, Amtrak's dual-powers don't operate west of Albany on the Lake Shore Limited, so a change of power is still involved.

Now, now, no one is going to tar and feather anyone on this thread!

Paul,

What is your opinion of the Fairbanks-Morse P-12-42...?

http://en.wikipedia.org/wiki/FM_P-12-42

It seems this FM product was superior to the FL9 in many ways - better fuel economy in diesel mode, higher max speed (117 mph vs 70 mph for the FL9) - but was overlooked (or undersold) vs the EMD product.

For what it's worth, I feel any locomotive that lasts 50 years should be given respect, despite the misgivings!

 

[MichaelSol]

... but what about the cost of power?  How many new power plants would be needed to power all U.S. railroads? 

None.

And what would this new demand do to the price of electricity for railroads and for general customers?

Probably lower it.

And what would be the cost comparison of sequestering all CO2 from all railroad diesels, versus sequestering it from the coal fired power plants producing power for electrified railroads?

Assuming, arguendo, that "CO2 sequestration" survives the current media hysteria and bizarre Supreme Court decisions, this is one factor that will act in favor of railroad electrification.

 

[Anonymous]

On page 10 of the July issue of TRAINS there's a brief news item regarding the possibility of reviving plans for electrification due to the high cost of diesel fuel.  The consensus conclusion of the author seems to be that mainline electrification is still too expensive even with today's fuel prices due to the upfront costs of stringing catenary and buying whole fleets of electric locomotives.

Suppose it were decided that all U.S. railroads would electrify.  There would be the cost issue of new catenary and new locomotives, but what about the cost of power?  How many new power plants would be needed to power all U.S. railroads?  And what would this new demand do to the price of electricity for railroads and for general customers?  And what would be the cost comparison of sequestering all CO2 from all railroad diesels, versus sequestering it from the coal fired power plants producing power for electrified railroads?

[beaulieu]

One point not mentioned in the article that needs to be brought up, is the effects of Tier III and Tier IV Emissions Standards when they come into force. They will likely have an impact on the economics of electrification vs. Diesel-Electric, by raising the cost of new Diesel-Electric Locomotives and probably their on-going maintenance costs. It looks like Diesel-Electrics will need to have Catalytic Convertors and probably Particle Filters which will require periodic maintenance, and replacement. Tier III is guaranteed to happen soon, and Tier IV is in the early stages of discussions concerning what is possible. Another point is that California may force the issue of railroad electification.