Electric braking-generative trains?

[CSSHEGEWISCH]

I was wondering one thing: does a diesel electric have any advantage over an electric loco when pulling heavy tonnage?

I wonder if the reason Euorpean railways are beginning to use diesel-electrics to supplement the overhead grids is becasue of a capacity issue of the grids.  Other threads have related how Euro-freights tend to be shorter and faster.  Heaver implies more power; is the system at capacity and need a boost, or is it a matter of getting the power down the cantenary.  The wires can only carry so much current without a significant voltage drop which would affect the available power.  By supplementing with diesel electrics, the trains that require teh extra power can carry it along with them...

This was one of the factors considered when New Haven ordered the FL9's.  The Cos Cob power plant was obsolescent at the time and NH was looking for ways to reduce the load on the plant.  In peak periods, the FL9's would primarily run west of New Haven to reduce the load on the power plants.

[MichaelSol]

Sort of suprrised that the M-G sets were that loud when starting and not at all surprised that they were quiet when running (been around a fair amount of electrical machinery). Thanks for sharing that tidbit.

Stopped at Gold Creek Substation on a tour to show the insides of a substation to a group, which happened to include David P. Morgan of Trains. I am sure he had been in a substation before, but the subject never came up.

After the tour of the machine room and the transformer room, everyone gathered in the office, with the door to the machine room open. Electrical Engineer George Frazier got on the mike and called the operator on duty at Janney and asked him start Motor Generator Set No. 1, the closest to the office.

A klaxon signaled the machine room to be clear, then the Control Data signal equipment near the door, inside the office, started receiving the sing-song code transmission -- a tone code -- which was very audible. As soon as the chattering relay kicked in, it was time to hold your ears with that door open like it was. When that relay stopped chattering, the MG set would start up. Most of the folks there didn't know that. So, everyone there is distracted by the big grey cabinet singing and chattering to itself.

Then the chattering stopped. There was a momentary silence as a series of relays aligned themselves.

Then "Bam." I think everybody jumped two feet, and then quickly stepped back from the door. It was a tremendous roar, an indescribable sound. If anyone said anything six inches away, you couldn't hear it.

When the machine reached synchronization after 15-20 seconds and quieted down, Morgan looked at me, and exclaimed "oh my goodness."

 

[erikem]

During college I took a number of electrical engineering courses, and one of my professors told me something interesting about the Commonwealth Edison Company.  It was the company's practice during the late '60s and early '70s to take a reading each night of the sweep second hand of a certain clock in downtown Chicago.  The clock was powered by a "hysteresous-synchronous motor," one whose speed was solely the function of a.c.-line frequency. 

Sometime between the hours of 2 and 3 a.m., the position of that second hand would be compared to the National Bureau of Standards WWV time signal.  If the second hand was running slow, Edison would order a slight increase in line voltage frequency.  If the clock was a little fast, the opposite would happen.  The adjustments were made through the late night hours so that by the time most people were waking up the next morning, the clocks throughout Chicagoland would be right on the money!  That's a nice little attention to detail, don't you think?    

Power companies have for decades tried to keep system time "on-time" - a typical spec is +/- 3 seconds. Back in the 1970's PG&E was using a digital clock that essentially counted the cycles and the system frequency would be tweaked to bring the system time back to NBS time. The governors on power plants are set up to have a slight variation in frequency with load - this provides damping during system transients. Most power companies have a way of telling the individual generating plants to vary the set-points on the governors - useful for noth frequency control and economic dispatch.

Curious where you took your EE courses - took mine at UC Bezerkeley in the mid-70's, including some power systems courses, a course in electrical machinery (which proved useful in working on NMR hardware) and a course in power electronics.

[Bob-Fryml]

One thing that does happen is that a regenerating traon can cause the line frequency to rise slightly, which then causes the governors on the power plant generators to back off a bit. Conversely, when a train is heading upgrade, the extra load causes a slight decrease in line frequency which is then compensated by the power plant governors to increase power a bit.

How incredibly obvious!  I never considered that the electric utility, by regulating the speed of its steam or hydropower turbines, could compensate for the outbound current being pumped out of a Milwaukee Road substation.  What a clean, elegantly simple explanation for something that's been puzzling me for I bet 40+ years.  Thanks, Erikem, for the "Eureka!" moment and please know that the gray matter between my ears feels a tad bit lighter right now!

= = = = = = = = = = = = = = = = = = = = = =

Here's a little story about being "in synch" that readers may find of interest. 

During college I took a number of electrical engineering courses, and one of my professors told me something interesting about the Commonwealth Edison Company.  It was the company's practice during the late '60s and early '70s to take a reading each night of the sweep second hand of a certain clock in downtown Chicago.  The clock was powered by a "hysteresous-synchronous motor," one whose speed was solely the function of a.c.-line frequency. 

Sometime between the hours of 2 and 3 a.m., the position of that second hand would be compared to the National Bureau of Standards WWV time signal.  If the second hand was running slow, Edison would order a slight increase in line voltage frequency.  If the clock was a little fast, the opposite would happen.  The adjustments were made through the late night hours so that by the time most people were waking up the next morning, the clocks throughout Chicagoland would be right on the money!  That's a nice little attention to detail, don't you think?    

[JSGreen]

I was wondering one thing: does a diesel electric have any advantage over an electric loco when pulling heavy tonnage?

I wonder if the reason Euorpean railways are beginning to use diesel-electrics to supplement the overhead grids is becasue of a capacity issue of the grids.  Other threads have related how Euro-freights tend to be shorter and faster.  Heaver implies more power; is the system at capacity and need a boost, or is it a matter of getting the power down the cantenary.  The wires can only carry so much current without a significant voltage drop which would affect the available power.  By supplementing with diesel electrics, the trains that require teh extra power can carry it along with them...

[Flint Hills Tex]

Bentnose Willie wrote:
"Long story short - European and North American railroad systems are very different from one another, and some of those differences play into why electrification isn't used on a large scale in North America."


I was wondering one thing: does a diesel electric have any advantage over an electric loco when pulling heavy tonnage? The reason I ask is, a number of railroads over here in Germany have been implementing diesel electrics from GE/Bombardier and EMD to haul heavy commodities such as coal, iron ore, and the like. That seems surprising, considering the high cost of diesel over here and the readily available electric power from the overhead caternary.

[erikem]

Each Motor Generator Set had small direct current generators mounted at the ends of the MG set shafts. On one end was a 12 kw capacity generator, and on the other end was a 30 kw capacity generator. The smaller unit provided field excitation for the two 1500 volt DC generators on the unit, and the 30 kw generator provided the field excitation for the 600 rpm synchronous AC motor.  The 30 kw generator output was "compounded" by the line current of the big DC generators. In essence, this controlled the power output of the AC motor, permitting it to either motor or regenerate to provide anything from 0 to 6000 kw ( 5 min. overload) without changing its speed or input/output voltage.

As pointed out, these big synchronous AC motors were determined to be stable at 60 cycles. That's how they were built. These things could be hit by lightning, and not change frequency.

They didn't like being asynchronous. When they were running asynchronous, during start up, those AC motors made a terrible roar. As soon as they synchronized, they ran remarkably quiet. You could be three miles away and know when a Milwaukee Road Substation was starting up its machines.

Most synchronous machines have amortisseur (spelling?) / damper windings - effectively the squirrel cage of a squirrel cage induction motor. In motors, these produce the torque necessary to get the motor going from topped to synchronous speed, in generators these windings helpp slow down the "rocking" after a system transient. Sort of suprrised that the M-G sets were that loud when starting and not at all surprised that they were quiet when running (been around a fair amount of electrical machinery). Thanks for sharing that tidbit.

Reminds me of visiting the Western Energy mine back in '71 (this was originally NP's Colstrip mine) - the loudest noise from the electrically driven shovel (27 cu. yd.) was the clang of the bottom of the bucket when opening.

[MichaelSol]

But regenerative braking is a whole nother animal.  If the kinetic energy of a descending train is being converted to electricity by the locomotive and that action raises the voltage of the trolley wire, when that higher voltage reaches the DC dynamos at a substation, won't that cause the motor generator set to spin faster?  And if the AC portion of the set is now in power generator mode, doesn't the faster spinning armature introduce spurious frequencies into the commercial power grid, frequencies that could be damaging to other electric motors using the same utility power grid?

I should think that the power being pumped out of a Milwaukee Road substation due to the movement of one or more trains in regenerative mode would have to be perfectly synchonized with the waveform of the commercial power grid supplying that substation.  How, with 1916 technology, was that accomplished?   

The M-G sets used synchronous motors. As long as torque on the shaft was less than pull-out torque, the motor would be running at exactly the speed set by the line frequency. An AC synchronous motor is not a lot different from an AC generator (alternator) and shaft speed remains locked to line frequency as long as the applied torque is less than pull-out. (Called pull-out because the motor or alternator is pulled out of synchronism).

Each Motor Generator Set had small direct current generators mounted at the ends of the MG set shafts. On one end was a 12 kw capacity generator, and on the other end was a 30 kw capacity generator. The smaller unit provided field excitation for the two 1500 volt DC generators on the unit, and the 30 kw generator provided the field excitation for the 600 rpm synchronous AC motor.  The 30 kw generator output was "compounded" by the line current of the big DC generators. In essence, this controlled the power output of the AC motor, permitting it to either motor or regenerate to provide anything from 0 to 6000 kw ( 5 min. overload) without changing its speed or input/output voltage.

As pointed out, these big synchronous AC motors were determined to be stable at 60 cycles. That's how they were built. These things could be hit by lightning, and not change frequency.

They didn't like being asynchronous. When they were running asynchronous, during start up, those AC motors made a terrible roar. As soon as they synchronized, they ran remarkably quiet. You could be three miles away and know when a Milwaukee Road Substation was starting up its machines.

[BentnoseWillie]

Capital investment and additional MOW costs are what I was thinking of. There's a large body of little things around those areas that make heavy electrification less suitable for North American applications than European ones.

Long story short - European and North American railroad systems are very different from one another, and some of those differences play into why electrification isn't used on a large scale in North America.

What host of large drawbacks? Alright, given, the capital investments which would be required to convert North American Railroads to electric power would be astronomical. Then there would be maintenance costs. But, living in Germany, I'm spoiled. I don't know any exact numbers, but the German Railway has invested quite an effort (and lots of capital) in getting as many miles of track under caternary as possible, and, yes, the new generation of electrics all have dynamic/regenerative braking capacity (they are what NJ Transit's new ALP 46 electrics are based on). Now why would they do something like that if it wasn't more economical? Your typical freight train over here doesn't come close to a North American train in length or tonnage, but you just need one loco, and, man, they sure move fast! When you have as many trains moving simultaneously as on Germany's railroads, it would be foolish not to utilize regenerative braking.

[arbfbe]

Bob-F,

 

When the substation operator knew he had a train about to go into regeneration he would reduce the line voltage so it was below the normal 3300-3500 volts.  That allowed a lower voltage in the trolley than what would be coming out of the traction motors so the current had a place with a lower potential to flow to.

The synchronous AC motors/generators in the substations only wanted to run at the frequency of the grid power.  They were amazingly self regualting in this matter.  If the frequency of the line was 60 cycle then the output from the MGs in the substation would be 60 cycles.

The differences between the various electrification systems are best summarized in the book "When the Steam Railroads Electrified" from Kalmbach.  It is long out of print and sometimes expensive on the used market but should be available in most city libraries.

[erikem]

But regenerative braking is a whole nother animal.  If the kinetic energy of a descending train is being converted to electricity by the locomotive and that action raises the voltage of the trolley wire, when that higher voltage reaches the DC dynamos at a substation, won't that cause the motor generator set to spin faster?  And if the AC portion of the set is now in power generator mode, doesn't the faster spinning armature introduce spurious frequencies into the commercial power grid, frequencies that could be damaging to other electric motors using the same utility power grid?

I should think that the power being pumped out of a Milwaukee Road substation due to the movement of one or more trains in regenerative mode would have to be perfectly synchonized with the waveform of the commercial power grid supplying that substation.  How, with 1916 technology, was that accomplished?   

The M-G sets used synchronous motors. As long as torque on the shaft was less than pull-out torque, the motor would be running at exactly the speed set by the line frequency. An AC synchronous motor is not a lot different from an AC generator (alternator) and shaft speed remains locked to line frequency as long as the applied torque is less than pull-out. (Called pull-out because the motor or alternator is pulled out of synchronism).

One thing that does happen is that a regenerating traon can cause the line frequency to rise slightly, which then causes the governors on the power plant generators to back off a bit. Conversely, when a train is heading upgrade, the extra load causes a slight decrease in line frequency which is then compensated by the power plant governors to increase power a bit.

There's also a bit of phase shift over the power lines as power is being transmitted - you can think of it as being equivalent to a long shaft - as you put more and more torque on the shaft, the more the shaft twists. Similarly, as you transmit more power over a transmission line, the more phase shift you see between the ends.

[Flint Hills Tex]

If you're thinking of this as a reason for more electrification of North American freight railroads, don't hold your breath. It's one small advantage compared to a host of large dreawbacks.

What host of large drawbacks? Alright, given, the capital investments which would be required to convert North American Railroads to electric power would be astronomical. Then there would be maintenance costs. But, living in Germany, I'm spoiled. I don't know any exact numbers, but the German Railway has invested quite an effort (and lots of capital) in getting as many miles of track under caternary as possible, and, yes, the new generation of electrics all have dynamic/regenerative braking capacity (they are what NJ Transit's new ALP 46 electrics are based on). Now why would they do something like that if it wasn't more economical? Your typical freight train over here doesn't come close to a North American train in length or tonnage, but you just need one loco, and, man, they sure move fast! When you have as many trains moving simultaneously as on Germany's railroads, it would be foolish not to utilize regenerative braking.

[Bob-Fryml]

This thread is really interesting because it's addressing an issue I've wondered about for decades.

Look at the MILW motor-generator substation model.  I can well understand what's going on:  an AC motor spinning more-or-less at a constant r.p.m. dictated by the frequency of the utility supplied current spins the DC dynamos that produce the trolley wire voltage.  If the frequency of the utility supplied current bumps up to 61, 62, or 63 hz., the DC output may rise a fraction of a volt, a couple of volts, or whatever because the dynamos are spinning faster.  If the frequency of the utility supplied current drops a little to 59, 58, or 57 hz., the DC output will drop a little.  Either way, it's no big deal with 3500-volt overhead.  The train still climbs the mountain with no perceptible change in speed.

But regenerative braking is a whole nother animal.  If the kinetic energy of a descending train is being converted to electricity by the locomotive and that action raises the voltage of the trolley wire, when that higher voltage reaches the DC dynamos at a substation, won't that cause the motor generator set to spin faster?  And if the AC portion of the set is now in power generator mode, doesn't the faster spinning armature introduce spurious frequencies into the commercial power grid, frequencies that could be damaging to other electric motors using the same utility power grid?

I should think that the power being pumped out of a Milwaukee Road substation due to the movement of one or more trains in regenerative mode would have to be perfectly synchonized with the waveform of the commercial power grid supplying that substation.  How, with 1916 technology, was that accomplished?   

[erikem]

MILW had looked at rectifiers, and several studies recommended their purchase to augment the system. About $325,000 each. Inverters were about the same cost, but MILW already had the rotating machinery which achieved the same result, so inverters were never considered for regeneration.

I was thinking more along the lines of BART and light rail lines with respect to inverters (specifically dual converters). The motor control circuits on the first generation of BART cars were set up to use regenerative braking if there was a load to take the regenerated power, otherwise dynamic braking resistors would kick in when the third rail voltage went above 1100 to 1200V (BART was nominally 1000V). One work-around would be putting in capacitor banks (ultra-caps are technically feasible, but not necessarily economically feasible) or flywheels (Pentadyne has some for data centers that float on ~500V DC busses - wouldn't be too much of a stretch to bump that up to 600V). Having the energy storage could smooth out the demand a bit.

The M-G sets on the MILW were just fine for regen - can't ask for a better load than a synchronous motor. There were several places that rectifiers could have been dropped in where the extra power was needed for motoring not regen.


In case anyone is wondering - a Dual Converter is noting more than a pair of three phase bridge rectifiers using thyristors (AKA SCR's) - where one bridge is hooked up in the same way as straight diodes and the other bridge hooked up in the opposite direction. By adjusting which bank of thyristors is being fired, the dual converter can transfer power from the AC to DC or from DC to AC.

[MichaelSol]

A slight quibble on the wording (your explanation is mostly  dead on) - the M-G sets were not reversed when the direction of power flow changed, the only thing that reversed was AC machine was leading the power grid in regeneration rather than lagging as in motoring (this can be seen by putting a AC-line synch'ed stroboscope on the shaft of the M-G set). Put it simply, motoring occurred when the DC bus voltage was less than the no-load voltage of the DC generators and regeneration occurred when the DC bus voltage was higher than the no load voltage of the DC generators.

Exactly right. Poor wording on my part -- the MG sets, the rotating machinery, rotated in only one direction; the power flow "reversed".

MILW had looked at rectifiers, and several studies recommended their purchase to augment the system. About $325,000 each. Inverters were about the same cost, but MILW already had the rotating machinery which achieved the same result, so inverters were never considered for regeneration.

[arbfbe]

The MILW system used a motor generator system in the substation which had a large AC motor with a DC generator on each end of the shaft.  When a train needed power the AC motor turned the tDC generators.  When a train was in regeneration then the locomotive traction motors provided power to turn the DC generators in the substations into motors which turned the AC motor into a generator which provided synchronous AC to the grid..  There was a combination of manual and automatic control used on the system which was designed near the turn of the centuries as well as a few system upgrades after.

The modern AC diesel locomotives from EMD and GE use alternators to generate AC power.  Solid state rectifiers then convert the AC power to a pure form of DC which is then fed through another set of solid state devices to provide the AC to power the traction motors.  The same system could be used today to convert the train supplied power to what ever choice of electricity the grid would need to accept.

I can just hear the power distribution desk now calling the railroad to ask if they could run some trains downhill since it is about 5:30 p.m. and peak usage is coming on.  Perhaps they could even have the railroads stage their coal trains bringing coal to the power plants at the tops of grades until additional capacity is needed for peak demand periods.  Now if you can just get permission to build a few long yard tracks at the tops of the hills and figure out how to run the trains up the hill during low demand, keep the crews from running out of time while they are waiting for peak time to come around and hold all up grade intermodal trains while it is a downhill railroad you have a place in railroad management.  

[erikem]

Milwaukee Road's early advertising about "white coal" referred to the use of water power exclusively at that time to supply Milwaukee's DC system. Milwaukee itself had surveyed and purchased several of the early hydroelectric sites, and participated in their development in order to supply the railway company. The 'battery" was the reservoir storage. If the MILW regenerated, the power company could use that power to supply other users rather than "use down" the water storage capacity to supply those users.  In essence, regeneration "saved" battery storage capacity.

MILW generated clean waveforms to the power companies by reversing the stationary Motor Generator sets at the Substations, producing 2300 v AC current from the 600 rpm synchronous AC motors, and stepping up to 110 kV AC through transformers. Most of the losses on regeneration are line losses -- AC, DC doesn't matter. The conversion losses, AC to different AC, or DC to AC were small by comparison.

A slight quibble on the wording (your explanation is mostly  dead on) - the M-G sets were not reversed when the direction of power flow changed, the only thing that reversed was AC machine was leading the power grid in regeneration rather than lagging as in motoring (this can be seen by putting a AC-line synch'ed stroboscope on the shaft of the M-G set). Put it simply, motoring occurred when the DC bus voltage was less than the no-load voltage of the DC generators and regeneration occurred when the DC bus voltage was higher than the no load voltage of the DC generators.

The GN single phase electrics and the VGN EL-2's had M-G sets in the cab, so regeneration worked in a manner very similar to the Milwaukee's system.

It isn't that difficult to set up a solid state "rectifier" substation to accept regenerated power - "dual converters" have been around for decades - and they were one of the subjects in a power electronics class I took slightly over thirty years ago. The waveforms aren't any worse than what a plain rectifier would impose on the line.

[MichaelSol]

I was under the impression that heavy railway electrification was not supplied by the regular distribution grid, but by dedicated sources.

It is supplied by the grid.

Until recently though it was difficult to get a clean enough waveform that the electricity companies would accept, traction circuits produce a lot of dirty harmonics. Power electronics changed all that, fortunately.

On Milwaukee, the power was in effect stored.

Milwaukee Road's early advertising about "white coal" referred to the use of water power exclusively at that time to supply Milwaukee's DC system. Milwaukee itself had surveyed and purchased several of the early hydroelectric sites, and participated in their development in order to supply the railway company. The 'battery" was the reservoir storage. If the MILW regenerated, the power company could use that power to supply other users rather than "use down" the water storage capacity to supply those users.  In essence, regeneration "saved" battery storage capacity.

MILW generated clean waveforms to the power companies by reversing the stationary Motor Generator sets at the Substations, producing 2300 v AC current from the 600 rpm synchronous AC motors, and stepping up to 110 kV AC through transformers. Most of the losses on regeneration are line losses -- AC, DC doesn't matter. The conversion losses, AC to different AC, or DC to AC were small by comparison.

[Hugh Jampton]

I was under the impression that heavy railway electrification was not supplied by the regular distribution grid, but by dedicated sources.

It is supplied by the grid.

Until recently though it was difficult to get a clean enough waveform that the electricity companies would accept, traction circuits produce a lot of dirty harmonics. Power electronics changed all that, fortunately.

[JSGreen]

capacitors (and batteries) would work for storing DC, but you would need a "Tuned Tank" resonant circuit, with a capacitor and a coil, to store AC.  I love to see that sucker!  But dont take any wrist watches or pacemakers near it...the magnetic field generated by a coil that big would be something to behold....

[dldance]

Current electrical storage research is focusing on very large capacitors which may add to the practicality of electric railroading.

dd

[JSGreen]

My point is that regenerated energy has to have somewhere to go - it cannot be stored with existing technologies.

Well, maybe it's picking a nit, but "Not Exactly". I would, however, agree that it can't be stored direcetly as electricity...(yet)

 There is a Nuke plant in Virginia I took some students to a few years ago(1991).  Instead of building a Nuke plant big enough to supply the peak demand, they built a damn and a lake.  During non-peak electrical use, instead of operating at a non-economical low output level, they generate extra power and pump water from the river up to the lake.  during Peak electrical useage, they use the water for Hydro generation.  Certainly there are losses involved, but they are storing the excess electrical energy for use later...with "just a few" conversion losses...

I believe the plant was North Anna, but the current web site doesent mention the "offline storage".  Perhaps it turned out not to be economical....but it was in fact physically possible!

[BentnoseWillie]

If there is not an ascending train to use the power, the power is supplied back to the power company supply and is metered for credit.

I was under the impression that heavy railway electrification was not supplied by the regular distribution grid, but by dedicated sources. If that's not the case then yes, energy could be returned to the grid, assuming the railway operates on AC, preferably at 60 Hz. If it's AC of another frequency, or DC, conversion losses would come into the equation.

On the Milwaukee Road, up to 50% of an ascending train's power needs could be supplied by an equivalent train regenerating on the downgrade.

Yes...as long as there were ascending and descending trains operating concurrently.

My point is that regenerated energy has to have somewhere to go - it cannot be stored with existing technologies.

[adrianspeeder]

Your idea is what is done.

Adrianspeeder

[MichaelSol]

Regenerative braking can work well, but it depends upon another train being under load at the same time, since storage of such amounts of energy is highly problematic.

If there is not an ascending train to use the power, the power is supplied back to the power company supply and is metered for credit. On the Milwaukee Road, up to 50% of an ascending train's power needs could be supplied by an equivalent train regenerating on the downgrade.

[BentnoseWillie]

Regenerative braking can work well, but it depends upon another train being under load at the same time, since storage of such amounts of energy is highly problematic.

I think subways already use this too, as they operate at a high enough density that there's almost always trains accelerating and decelerating at the same time.

If you're thinking of this as a reason for more electrification of North American freight railroads, don't hold your breath. It's one small advantage compared to a host of large dreawbacks.

[erikem]

If trains were powered by an external electric grid, and these same trains has the capability to generate electricty as a part of the braking process, would certain RRs that have lots of grades in their network benefit from this? As a train is going 'down hill' it generates power that is fed back into the grid, and this power then supplements the power required for a train going uphill. For some RRs this could provide decent savings for their energy needs. Yes maybe no?

The Milwaukee estimated that they recovered ~15% of the electricity used to move trains uphill. Not a huge savings, but nothing to sneeze at either. As you hinted, this was especially welcome when a descending train was meeting an ascending train.

Regeneration was also used by the N&W, Virginian and Great Northern, though the last set of locomotives for the Virginian did not support regeneration. The cars used on BART and most light rail lines are capable of regenerative braking - the limitation is in the power supply.

[spokyone]

http://www.scn.org/cedar_butte/milw-elec.html

 

An excerpt from the page about the Milwaukee road.

Perhaps the coolest aspect of the electrification system adopted by the Milwaukee was regeneration. When trains went down hill the electric motors were used as generators. This both slowed the train down (with substantial savings in the cost of replacing brake-shoes), and returned power to the system to help ascending trains, reducing the overall power needs by about twelve percent. Heavy capital costs, but significant operational savings (despite padding of costs); the net benefit was later found to amount to a return on investment of nine percent annually.

[CrazyDiamond]

If trains were powered by an external electric grid, and these same trains has the capability to generate electricty as a part of the braking process, would certain RRs that have lots of grades in their network benefit from this? As a train is going 'down hill' it generates power that is fed back into the grid, and this power then supplements the power required for a train going uphill. For some RRs this could provide decent savings for their energy needs.

Yes maybe no?