Dynamic brakes on locomotive hauled trains - technical help wanted

[GC class B1]

I am pretty familiar with pneumatic brake systems on rolling stock, but I only know the basics about dynamic braking. I am hoping that some of the knowledgable people on the forum can explain in more detail how dynamic brakes interface with friction braking. I understand that for class 800 the when the dynamic brake is enabled, the friction brake does not apply unless wheel slide is detected or the dynamic brake does not operate as designed. In the event that the dynamic brake does not operate the friction brake will apply. As the class 800 brake is electronically controlled it would seem that both the dynamic and friction brake system blending can be achieved quite easily.
I would like to know how the dynamic brake functions on a locomotive hauled train operate and brake blending is achieved. From my limited knowledge I know that the Driver’s Brake Controller controls both the friction and dynamic brake on the locomotive and the friction brake on the vehicles being hauled. Could someone please answer the following questions: 1. Does the dynamic brake only operate on the locomotive and the friction brake on the hauled vehicles always apply, or do the friction brakes on the hauled vehicles only apply when the locomotive dynamic brake is not providing enough brake force. The maximum benefit from dynamic braking will be achieved if the dynamic brake provides all the brake force for the train but this requires a higher wheel/rail coefficient of friction than if the locomotive dynamic brake is only braking the weight of the locomotive. The dynamic brake working on its own will significantly increase the likelihood of wheel slide occurring in lower adhesion conditions.
2. How does the Locomotive dynamic brake apply? Does the dynamic brake respond to the reduction in Brake Pipe pressure or is the Brake Pipe pressure controlled by the locomotive’s dynamic brake control. I am wondering how the Brake Pipe pressure is controlled during blended braking.

 

[Fragezeichnen]

Whether and how blending functions on a locomotive depends on how the controls on that locomotive have been designed.

1. The vast majority of German locomotives, and more recent UK locomotives such as the Stadler UKDual have 2 brake controllers, not one(or 3 if you count the loco-only friction brake!). The driver must decide if the dynamic brake alone will be "enough". Since on a long freight train applying and releasing the train air brakes may take a long time, it is important to give the driver the choice to avoid using them if they want to(e.g. to hold the speed steady on a descending gradient). Older UK locomotives(e.g. class 86) have no seperate dynamic brake lever. I am not sure if they either have no dynamic brake, of if it is not possible to operate it alone, or if there is some control scheme like the first notch being dynamic-only.

2. Since the locomotive cannot know the braking properties of the train it is hauling, the driver must always select the exact brake pipe pressure they want and the brake control system cannot interfere with this. As for combining that with dynamic braking, on older German locomotives the driver must operate both brake controllers simultaneously to apply both types of brakes, and dynamic brake lever can be optionally mechanically connected to the train brake lever to make this easier. In newer locomotives the traction control computer will automatically apply blended dynamic brake when the train brake is operated.

 

[hexagon789]

Older UK locomotives(e.g. class 86) have no seperate dynamic brake lever. I am not sure if they either have no dynamic brake, of if it is not possible to operate it alone, or if there is some control scheme like the first notch being dynamic-only.

86s and 87s will use rheostatic braking at 20mph+ instead of the tread brakes rather than blending them. A 5psi application is maintained to kept the brake rigging taught and reduce the time for it to switch back in as the speed reduces or if it cuts out for any reason.

The main reason for using it is thus to reduce brake block wear.

Rheo isn't available if going through a neutral section or if any traction motors are isolated.

This contrasts to 90s where the basic idea is the same but rheostatic braking can be maintained through neutral sections and is still available if one motor is isolated albeit giving reduced brakeforce.

85s are the same as 86/87, but originally had a separate rheostatic brake function. The driver would move a lever to switch from motoring to rheostatic braking and then use the tap-changer handle to control the rheostatic brakes. Unfortunately it was found temperamental in use and also some drivers were too heavy handed - it relied on manually watching the current and adjusting according through the speed range.

So it was altered to automatic operation, slaved to the train brake and the 86s and 87s set-up accordingly.

With all these classes, rheostatic braking is only available with use of the train brake, application of the locomotive brake does not cause a rheostatic application.

 

[GC class B1]

86s and 87s will use rheostatic braking at 20mph+ instead of the tread brakes rather than blending them. A 5psi application is maintained to kept the brake rigging taught and reduce the time for it to switch back in as the speed reduces or if it cuts out for any reason.

The main reason for using it is thus to reduce brake block wear.

Rheo isn't available if going through a neutral section or if any traction motors are isolated.

This contrasts to 90s where the basic idea is the same but rheostatic braking can be maintained through neutral sections and is still available if one motor is isolated albeit giving reduced brakeforce.

85s are the same as 86/87, but originally had a separate rheostatic brake function. The driver would move a lever to switch from motoring to rheostatic braking and then use the tap-changer handle to control the rheostatic brakes. Unfortunately it was found temperamental in use and also some drivers were too heavy handed - it relied on manually watching the current and adjusting according through the speed range.

So it was altered to automatic operation, slaved to the train brake and the 86s and 87s set-up accordingly.

With all these classes, rheostatic braking is only available with use of the train brake, application of the locomotive brake does not cause a rheostatic application.

Thank you for this detailed explanation. If I understand correctly, the answer to my first question is that the friction brake on the hauled vehicles will always apply whenever the driver makes a brake demand.

 

[hexagon789]

Thank you for this detailed explanation. If I understand correctly, the answer to my first question is that the friction brake on the hauled vehicles will always apply whenever the driver makes a brake demand.

Yes, when hauled by an 85, 86, 87 or indeed 90, the friction brakes on the hauled vehicles will always apply when a train brake application is made. The rheostatic brake purely reduces brake wear on the locomotive, it doesn't provide a braking effort on it's own.

The even older DC 76s and 77s were different; not only did they have regenerative as well as rheostatic braking but both were manually controlled and those braking systems could be and were used to control train speed alone without any friction brake application.

 

[edwin_m]

Repeated application and partial release of a single-pipe air brake will eventually drain the reservoirs and make the train brakes ineffective. The dynamic brake is used on long descents to avoid this happening - the source I read was North American but I assume the same is true in Europe too. But I guess for that to be done safely, the driver must have independent control of the two brakes so as to select the appropriate one.

 

[hexagon789]

Repeated application and partial release of a single-pipe air brake will eventually drain the reservoirs and make the train brakes ineffective. The dynamic brake is used on long descents to avoid this happening - the source I read was North American but I assume the same is true in Europe too. But I guess for that to be done safely, the driver must have independent control of the two brakes so as to select the appropriate one.

With triple valves yes, but using distributors instead gets round this.

Also passenger trains (including in the US) are almost universally twin-pipe these days.

Before dynamic brakes, ie in the steam or early diesel eras, retainers were used for long descents.

Some US roads still have them as a back-up I understand.

 

[Bill57p9]

Thank you for this detailed explanation. If I understand correctly, the answer to my first question is that the friction brake on the hauled vehicles will always apply whenever the driver makes a brake demand.

Just for clarity: most if not all UK locos have a locomotive brake, which operates the friction brakes on the locomotive only. Therefore it is possible for the driver to make a locomotive friction brake demand without any brake demand on the hauled vehicles.
That said, using the loco brake with a train in motion would be incorrect technique as far as I am aware.

There is a good explanation of the braking system on the class 92 (dynamic & friction) and mk5s in the RAIB report into the Caledonian Sleeper runaway from page 16 onwards.

 

[GC class B1]

Just for clarity: most if not all UK locos have a locomotive brake, which operates the friction brakes on the locomotive only. Therefore it is possible for the driver to make a locomotive friction brake demand without any brake demand on the hauled vehicles.
That said, using the loco brake with a train in motion would be incorrect technique as far as I am aware.

There is a good explanation of the braking system on the class 92 (dynamic & friction) and mk5s in the RAIB report into the Caledonian Sleeper runaway from page 16 onwards.

Thank you. I have now read this report and it is has now raised an uncertainty in my mind as to how dynamic brake on class 92 operates. On post No. 3, hexagon789 provided an answer to my question No.1 in my original post and this appears to be confirmed In paragraph 31 of the report where it is explained that the Rheostatic brakes replaces the locomotive friction brake to reduce locomotive brake block wear. In paragraph 83 of the report however it is explained that the entire train was being retarded by the locomotive rheostatic brake. For the locomotive rheostatic brake to mask the non application of the friction brakes on the coaches as stated in this paragraph the rheostatic brake would need to provide more than three times the locomotive friction braking force. This would require a very high wheel/rail coefficient of friction and contradicts the statement in paragraph 31. Have I misunderstood the content of the report you quoted.

 

[hexagon789]

Thank you. I have now read this report and it is has now raised an uncertainty in my mind as to how dynamic brake on class 92 operates. On post No. 3, hexagon789 provided an answer to my question No.1 in my original post and this appears to be confirmed In paragraph 31 of the report where it is explained that the Rheostatic brakes replaces the locomotive friction brake to reduce locomotive brake block wear. In paragraph 83 of the report however it is explained that the entire train was being retarded by the locomotive rheostatic brake. For the locomotive rheostatic brake to mask the non application of the friction brakes on the coaches as stated in this paragraph the rheostatic brake would need to provide more than three times the locomotive friction braking force. This would require a very high wheel/rail coefficient of friction and contradicts the statement in paragraph 31. Have I misunderstood the content of the report you quoted.

I believe the 92s operate differently and have a much stronger rheostatic brake which can be used irrespective of any other brake application.

I believe they follow European practice, by having a PBL air brake where you add or subtract air as you wish and the controller springs to a 'hold' position; as opposed to the traditional British self-lapping design.

 

[ac6000cw]

Before dynamic brakes, ie in the steam or early diesel eras, retainers were used for long descents.

Some US roads still have them as a back-up I understand.

In the US, the EMC/EMD 'FT' locos from 1940 onwards (their first mainline freight diesels) offered dynamic braking as an option, so it goes back a long way in the diesel era (and much further back with electric locos). It pretty much revolutionized the handling of heavy freight trains in mountainous areas. It's been standard equipment on new US locos for quite a long time, and there are some long, steep descents where trains are not allowed to descend without sufficient dynamic brake capability for safety reasons (I think Cajon Pass is one of them - the steepest of the three BNSF tracks there is a 3% gradient in places).

Re. 'retainers' on US freight cars (for those unfamiliar with the term) - when they are 'turned up' via a handle on each vehicle a certain level of air pressure in the brake cylinders is retained (and hence braking effort) after a brake application and release. They would normally be 'turned up' at the top of a steep descent - the equivalent of 'pinning down' brakes on unfitted wagons in the UK back in the day.

AFAIK US freight cars are still equipped with retainers, but with modern high-capacity dynamic braking on locos there are probably not that many situations where they are regularly used. I remember reading some years ago that Southern Pacific used them when taking heavy iron-ore trains down the 3% gradients on the ex-DRGW Tennessee Pass route in Colorado (which is now closed as a through route - the operating difficulties being one reason).

 

[hexagon789]

In the US, the EMC/EMD 'FT' locos from 1940 onwards (their first mainline freight diesels) offered dynamic braking as an option, so it goes back a long way in the diesel era (and much further back with electric locos). It pretty much revolutionized the handling of heavy freight trains in mountainous areas. It's been standard equipment on new US locos for quite a long time, and there are some long, steep descents where trains are not allowed to descend without sufficient dynamic brake capability for safety reasons (I think Cajon Pass is one of them - the steepest of the three BNSF tracks there is a 3% gradient in places).

By early diesel I was thinking more 1930s - ie prewar.

But even when they became common, some roads banned their use.

When you look at some of the rules some roads had, you do wonder why.

Take the Pennsy for instance, they never fitted graduated release to their passenger cars. Traditionally engineers would power brake to achieve smooth and precise stops. The Pennsy? Also banned power braking. Instead engineers were to reduce speed to 10mph, release the brakes and then bring the train to a stop using a second lesser application.

Re. 'retainers' on US freight cars (for those unfamiliar with the term) - when they are 'turned up' via a handle on each vehicle a certain level of air pressure in the brake cylinders is retained (and hence braking effort) after a brake application and release. They would normally be 'turned up' at the top of a steep descent - the equivalent of 'pinning down' brakes on unfitted wagons in the UK back in the day.

Modern ones have three settings:

Release - does nothing, brakes behave normally
High - holds minimum 20psi application
Slow direct - brakes release more slowly

I believe historically Low and Medium settings of various 'retained' brake cylinder pressures also existed.

AFAIK US freight cars are still equipped with retainers, but with modern high-capacity dynamic braking on locos there are probably not that many situations where they are regularly used. I remember reading some years ago that Southern Pacific used them when taking heavy iron-ore trains down the 3% gradients on the ex-DRGW Tennessee Pass route in Colorado (which is now closed as a through route - the operating difficulties being one reason).

I know some are still equipped, but with the advent of not just dynamic braking but long range dynamic braking with AC traction motors that can almost completely stop a train on their own (I believe full effectiveness to 8mph is typical and some braking effort to as low as 3 or 5mph is possible), I thought they were no longer universal.

 

[ac6000cw]

I believe historically Low and Medium settings of various 'retained' brake cylinder pressures also existed.

From what I've read, historically they usually had Release/High(20psi)/Low(10psi) settings. In the early 1950s Westinghouse added 'slow direct release' setting as a fourth setting, then the 'Low' setting was replaced by the 'slow direct release' setting in the mid-1960s to get the modern three position 'Release/High/Slow Direct release' retainer (as you mention).

I know some are still equipped, but with the advent of not just dynamic braking but long range dynamic braking with AC traction motors that can almost completely stop a train on their own (I believe full effectiveness to 8mph is typical and some braking effort to as low as 3 or 5mph is possible),

My understanding re. extended-range dynamic braking minimum speeds is the same as yours.

I thought they were no longer universal.

I don't have any inside knowledge, but as far as know/can find I believe retainers are still fitted e.g. the Canadian TSB accident report into the 2019 fatal runaway accident on Kicking Horse Pass makes reference to setting retainers: "In order to limit the train’s acceleration after the brakes were released, the pressure retaining valves had to be set to the high-pressure position on 84 of the rail cars." (on page 8).

 

[Harbon 1]

1. The vast majority of German locomotives, and more recent UK locomotives such as the Stadler UKDual have 2 brake controllers, not one(or 3 if you count the loco-only friction brake!). The driver must decide if the dynamic brake alone will be "enough". Since on a long freight train applying and releasing the train air brakes may take a long time, it is important to give the driver the choice to avoid using them if they want to(e.g. to hold the speed steady on a descending gradient). Older UK locomotives(e.g. class 86) have no seperate dynamic brake lever. I am not sure if they either have no dynamic brake, of if it is not possible to operate it alone, or if there is some control scheme like the first notch being dynamic-only.

Interestingly the move with the standard european cabs has been back to a single brake controller for both dynamic and train brake. I'm not sure when the change actually took place but the newer Traxx 2s (such as the 146.3s) which were introduced I think around 2011/12 had this design, and I believe subsequently the 68s do as well?

What changed I'm not sure since as you say the controllers were mechanically linked when applying the train brake, but you could use the dynamic on its own. Now it's just one lever and I assume the computer decides for the loco what is best. How that works on freight trains or on descents I'm not too sure!

Edit: as @bahnause points out the dynamic brake is now intergrated with the power handle rather than the brake handle!

 

[GC class B1]

Interestingly the move with the standard european cabs has been back to a single brake controller for both dynamic and train brake. I'm not sure when the change actually took place but the newer Traxx 2s (such as the 146.3s) which were introduced I think around 2011/12 had this design, and I believe subsequently the 68s do as well?

What changed I'm not sure since as you say the controllers were mechanically linked when applying the train brake, but you could use the dynamic on its own. Now it's just one lever and I assume the computer decides for the loco what is best. How that works on freight trains or on descents I'm not too sure!

Thank you for this interesting point. I can understand that class 800 would only need one brake controller for both the dynamic and friction brake as the electronic control could relate brake demand to the required deceleration rate and blend the brake accordingly. I would not expect this to work for locomotive hauled trains as the consist will vary and the deceleration rate achieved for a given brake controller setting will depend on the type of vehicle being hauled. The maximum deceleration rate for passenger trains will be higher than for freight trains.

 

[bahnause]

What changed I'm not sure since as you say the controllers were mechanically linked when applying the train brake, but you could use the dynamic on its own. Now it's just one lever and I assume the computer decides for the loco what is best. How that works on freight trains or on descents I'm not too sure!

It depends what the customer orders. Even with the newest TRAXX like the 147, the dynamic brake can still be operated separately. The dynamic brake is now integrated in the power controller on the left hand side. The train brake is on the right now, no need for the separate dynamic brake anymore.

 

[Harbon 1]

It depends what the customer orders. Even with the newest TRAXX like the 147, the dynamic brake can still be operated separately. The dynamic brake is now integrated in the power controller on the left hand side. The train brake is on the right now, no need for the separate dynamic brake anymore.

Ah that sounds like I've got slightly confused there then, having it on the power handle makes much more sense!

 

[ac6000cw]

North American diesel locos have sometimes had a single power/dynamic brake (DB) control with a power/DB selector switch, sometimes a combined control which is power in one direction and DB in the other (from the centre) or sometimes separate power and DB controls - it's varied over time. But I believe the power/DB control is always separate to the train brake control and the independent loco (friction) brake control.

Re. minimum speeds for dynamic braking on North American diesel locos, I came across this interesting post from user 'cadmus' on the Trainorders.com forum back in 2010 (note 'AC locomotive' means one with AC traction motors):

AC locomotives have what is called "nero zero" dynamic braking. If the DB controller handle is in position 8 (full DB)m for example, an AC locomotive (AC4400CW, ES44AC, SD9043AC, SD70MAC, SD70ACe) should produce maximum DB effort (as much as `115,000 pounds on an ACe) down to a speed of about 0.3-to-0.5 MPH and then drop linearly (straight line) to zero pounds of retarding force at zero MPH.

With the best DC locomotive with "extended range" and high-capacity DB, like an SD70M-2 or ES44DC, the braking or retarding effort will be around maximum (about 80,000 pounds braking effort) down to a speed of about 9-to-10 MPH and then drop linearly down to zero force at zero 0 MPH.

The "near zero" DB is an exceptionally valuable feature for an engineer having to stop a train on a grade. In some circumstances you can bring a train to near-stop without touching the air. When starting on a downhill, AC dynamics with near-zero also come in very handy ... remember, once you release the air brakes on a freight train, the brake shoes go "off" so to speak and the air compressors are forcing air back through the brake pipe to recharge the reservoirs on the cars. If you're launching off on a heavy grade, you pretty well have to be committed to "go" because when you kick off the brakes, the air system is recharging and you essentially don't have much opportunity to re-apply the air brakes until the reservoirs are recharged. (Releasing the brakes on a freight train from a downhill stop is not like moving your car and simply stepping on the brake pedal if you change your mind and want to stop again.)

AC dynamic braking also provides another important feature called "roll-back protection". When starting a train in power on a hill, if the train starts to roll backwards, the AC locomotive can be configured to provide "roll-back protection" in which the motors start to generate DB retarding effort until enough power throttle is set to make tractive effort greater than the rolling resistance and to overcome the gravity load. With a DC locomotive, if you don't put enough juice to the motors and you start to roll backwards, all you can do is apply more power or set the brakes.

Could "near zero" be implemented on DC locomotives? Not really. You can add additional contactors in the DB system of an SD40-2, for example, so that more DB effort can be squeezed out (by reducing the DB resistor grid resistance as speed drops) but you will never get true "near zero" DB as found on an AC locomotive.

For those interested in what's involved in getting heavy freight trains down long steep gradients, I recommend a read of the 'Getting them down the grade' article in US Trains magazine April 2004 'Mountain Railroads' issue. It's available from their online archive but to access it you need to be a subscriber or take advantage of the 'trains.com Unlimited' 30-day free trial subscription (which gives access to their entire magazine archive going back to 1940!).

 

[AM9]

I understand that dynamic braking is placing a load on the motors forcing them to generate power and thereby acting as a braking force. I also understand that the resultant electric power generated can be dissipated in two main areas:

1) by banks of load resistances on board the loco (rheostatic braking), - the only option for diesel-electric locos​

2) by converting the electrical power to line (usually OLE) voltage for use by other trains in the electrical section or even fed back into the primary supply network (regenerative braking). This is becoming standard on EMUs.​

In the posts above much mention is made of rheostatic braking but there is also mention of ac locos e.g. 86/87/90/92. Do these have the ability to feed power back onto the OLE and if so, would that be used on heavy freight applications of dynamic braking?

 

[hexagon789]

In the posts above much mention is made of rheostatic braking but there is also mention of ac locos e.g. 86/87/90/92. Do these have the ability to feed power back onto the OLE and if so, would that be used on heavy freight applications of dynamic braking?

85s, 86s, 87s and 90s are rheostatic only; there is no capability to recuperate braking energy and pass it back to the OLE.

92s do have regenerative capability but it is disabled; they can use rheo down to about 3mph.

As I mentioned above the electric braking on 85s, 86s, 87s, 90s simply replaces the braking effort on the locomotive it is not a dynamic brake in that it cannot be used to control the trains speed alone. It only operates under a train brake application anyway.

 

[edwin_m]

Regenerative braking on an AC supply requires some sophisticated electronics to ensure that the AC being pushed back into the line is exactly the same frequency and phase as what is there already. More recent locos and EMUs with AC motors rely on similar electronics to drive their motors at variable frequency, so they will generally also be able to regenerate, but older AC locos/EMUs with DC motors generally don't. That goes for all of classes 81 to 91.

Regenerating into a DC supply is much easier - the Woodhead locos did it with 1940s technology.

 

[AM9]

85s, 86s, 87s and 90s are rheostatic only; there is no capability to recuperate braking energy and pass it back to the OLE.

92s do have regenerative capability but it is disabled; they can use rheo down to about 3mph.

As I mentioned above the electric braking on 85s, 86s, 87s, 90s simply replaces the braking effort on the locomotive it is not a dynamic brake in that it cannot be used to control the trains speed alone. It only operates under a train brake application anyway.

Is there a particular reason why the '92s have their regen to ac disabled, - I presume that on DC it could be difficult to handle as specific provision would be needed in the case of sections with little else running on them.

 

[Poppysdad]

Its a long time since I worked on Loco brake systems but if I remember correctly on a Class 92 and Class 90 the brake pipe control unit either DW (Position dependent) or NW (Time dependent) reduces the Brake Pipe pressure. The loco distributors then respond to the drop in BP but the Brake Cylinder pressure output goes through a blending relay valve. There is a second input to the blending relay valve which is fed from a Electro Pneumatic convertor which is controlled by the traction system in proportion to the amount of Dynamic Brake being produced by the Traction system. The pressure from the EP convertor then backs off the distributor Brake cylinder pressure thus the brakes on the Loco are blended pneumatically with the BP on the train being controlled via the drop in brake pipe.

Brakes on trains like the Class 800 are controlled differently with the brake electronics initiating a friction brake application whilst simultaneously sending a Dynamic Brake demand signal to the traction system. The traction system then sends back a signal as to the level of Dynamic brake being achieved and the Brake electronics then reduces the friction brake demand in proportion to the amount of dynamic brake being achieved. Effectively the brakes are blended in the electronics.

 

[ac6000cw]

I also understand that the resultant electric power generated can be dissipated in two main areas:
1) by banks of load resistances on board the loco (rheostatic braking), - the only option for diesel-electric locos2) by converting the electrical power to line (usually OLE) voltage for use by other trains in the electrical section or even fed back into the primary supply network (regenerative braking). This is becoming standard on EMUs.

You could add storing recovered energy in batteries to that list e.g. the 756s and upcoming 93s.

 

[Taunton]

I presume that on DC it could be difficult to handle as specific provision would be needed in the case of sections with little else running on them.

This is handled by having resistances outside the substation to "burn off" electric power from the overhead when there was no other train in the section. Woodhead substations had this, along with various overseas installations. Regen was long confined to DC systems only, which have much shorter power sections and substation intervals than 25kV AC.

Contrary to much belief, when a train is regenerating back into the line it is by no means assured that there is another train, within the substation section, which is motoring and can absorb the power - in fact, something less than 50% of the time. I do believe that it was looked at that when regen was being "wasted" on the Woodhead the current might be used for an electric pump to pump water up between two of the reservoirs alongside the line, and later when power was demanded the same pump could act as a dynamo and generate current by running water back down again - nothing official' I think it was a university student study.

 

[hexagon789]

Is there a particular reason why the '92s have their regen to ac disabled, - I presume that on DC it could be difficult to handle as specific provision would be needed in the case of sections with little else running on them.

I assume it was why many classes of the same period that were equipped also had other disabled - inadequate power supply setup to take it.

 

[Poppysdad]

This is handled by having resistances outside the substation to "burn off" electric power from the overhead when there was no other train in the section. Woodhead substations had this, along with various overseas installations. Regen was long confined to DC systems only, which have much shorter power sections and substation intervals than 25kV AC.

Contrary to much belief, when a train is regenerating back into the line it is by no means assured that there is another train, within the substation section, which is motoring and can absorb the power - in fact, something less than 50% of the time. I do believe that it was looked at that when regen was being "wasted" on the Woodhead the current might be used for an electric pump to pump water up between two of the reservoirs alongside the line, and later when power was demanded the same pump could act as a dynamo and generate current by running water back down again - nothing official' I think it was a university student study.

Many years ago London Underground looked at the use of Flywheels to store energy , both on train and off train

 

[AM9]

This is handled by having resistances outside the substation to "burn off" electric power from the overhead when there was no other train in the section. Woodhead substations had this, along with various overseas installations. Regen was long confined to DC systems only, which have much shorter power sections and substation intervals than 25kV AC.

Contrary to much belief, when a train is regenerating back into the line it is by no means assured that there is another train, within the substation section, which is motoring and can absorb the power - in fact, something less than 50% of the time. I do believe that it was looked at that when regen was being "wasted" on the Woodhead the current might be used for an electric pump to pump water up between two of the reservoirs alongside the line, and later when power was demanded the same pump could act as a dynamo and generate current by running water back down again - nothing official' I think it was a university student study.

I was aware of the need for 'emergency' loads in which to dump power, - especially on DC lines, but my comment was more to do with ac where sections are generally much longer and thus more likely to have other trains drawing power, plus the fact that it is easier to send power back through a transformer, - particularly as a reverse feed isn't really a higher voltage but more an advanced phase feed.

 

[edwin_m]

Contrary to much belief, when a train is regenerating back into the line it is by no means assured that there is another train, within the substation section, which is motoring and can absorb the power - in fact, something less than 50% of the time. I do believe that it was looked at that when regen was being "wasted" on the Woodhead the current might be used for an electric pump to pump water up between two of the reservoirs alongside the line, and later when power was demanded the same pump could act as a dynamo and generate current by running water back down again - nothing official' I think it was a university student study.

This will be worse under DC. At the higher current and lower voltage, over the same distance the current flowing to another train generates much more voltage drop (absolutely and even more so as a proportion of line voltage) than it would be under a higher-voltage AC system. Therefore, it's much more likely that the voltage at the pantograph of the regenerating train will exceed a maximum limit. This effect is likely to outweigh the effect of sectioning the AC supply which would prevent AC regeneration into a train that's being supplied from a different phase or substation. On a DC supply all substations are connected in parallel.