Trains.com

Dynamic braking on subway trains

4731 views
18 replies
1 rating 2 rating 3 rating 4 rating 5 rating
  • Member since
    October 2020
  • 1 posts
Dynamic braking on subway trains
Posted by subwayguy on Friday, October 30, 2020 11:01 AM
I am looking for people with knowledge of subway train braking systems.
I think subways typically use what is referred to as ‘dynamic braking’. As I understand it, the dynamic braking system involves converting a train’s kinetic energy into more electric potential while also conserving brakes shoes. First, electrical brakes are used, then, to bring a train to full stop, the friction brakes are used. I’ve read that the system in Montreal Canada switches from electric to its friction brakes at a speed of 10km/h. I’ve also read that the Barcelona subway system switches at 5km/h.
 
 I'm looking for this value for Toronto's trains? The system uses the Bombardier A35T08 (Toronto Rocket) and Bombardier T1 trains. 
  • Member since
    January 2019
  • 951 posts
Posted by Erik_Mag on Friday, October 30, 2020 10:38 PM

The first rapid transit system that I know of to use dynamic/regenerative braking was BART, although subway cars using PCC controllers (think Chicago) may have implemented dynamic braking. Dynamic/regenerative braking is relatively easy to implement with solid state motor controls (e.g. chopper or inverter), so most subway cars built since 1970 are likely to have it.

Both dynamic and regenerative braking involve turning the motors into generators. The difference is that in dynamic braking, the generated electric power is dissipated in resistance grids, where in regenarive braking the generated power is returned to the traction power source (third rail for subways). BART implemented a blended system where the train "tries" to return power to the third rail for use by another train but will switch in resistors if there is no other train to accept the regenerated power.

  • Member since
    June 2002
  • 17,582 posts
Posted by daveklepper on Saturday, October 31, 2020 1:07 PM

All PCC cars used dynamic braking, including CTA's "Spam Cans (6000s).  More modern equipment generally uses regerative braking or a combination of regen. and dynamic.  Often, as on all pre-WWII PCCs, this is in combination with air, and streetcars usually have magnetic track brakes for emergency stops.       Rapid transit cars, even PCCs, usually do not.

By PCC, I refer to trucks, motors, control system, not the PCC body style used on streetcars.

  • Member since
    January 2019
  • 951 posts
Posted by Erik_Mag on Saturday, October 31, 2020 11:26 PM

Dave,

Thanks for jumping in with the PCC information - and I figured that your use of PCC referred to trucks, motors and control systems. I was guessing that the Chicago PCC subway cars might have been capable of dynamic braking, but I knew for sure that the BART cars had that capability.

  • Member since
    December 2007
  • From: Georgia USA SW of Atlanta
  • 10,242 posts
Posted by blue streak 1 on Sunday, November 1, 2020 5:23 AM

For dynamic braking or regeneration?  for the systems an example using thee NEC will suffice.  The CAT voltage is 12,000 volts nominal.  that is 10 % + / -.  That is 13,200 as plus and 10,800 minus. If an ACS-64 notes voltage is near that value or below it maintains regeneation that other equipment draws.     If the voltage goes near the upper limit the ACS=64 recognizes no other draw then it switches to dynamic.

If national power service is set up it will take the higher voltage and put power back into the national grid.  So all trains on the system will use reneration.  if the substations are not set up to return power to national grid then regeneration has to power other train(s) .  If not then dynamic braking is called for or trains will have to use normal wheel baking..

For subways and light rail that uses DC more often the same can be said however the return may be more difficult.  The Milwaukee DC used motor generators at substations that put regeneration power back to the grid.  If DC power systems do not use motor generators then the substations need inverters to put regeneration power back into the grid.   

 

  • Member since
    September 2011
  • 5,192 posts
Posted by MidlandMike on Monday, November 2, 2020 8:38 PM

AFAIK the NEC cat south of NY is still 25 Hz, so regeneration back into the 60 Hz grid would need that phase conversion.

  • Member since
    January 2019
  • 951 posts
Posted by Erik_Mag on Monday, November 2, 2020 10:20 PM

A fun fact that is little known by people who have had no education in electric power systems (a lot of EE's don't know this): power transfer on AC systems is more dependent on the relative phase of the generator with respect to the load than it is to the relative voltage. Voltage difference controls the VAR flow (VAR = Volt-Amperes Reactive).

As an example, the original LADWP transmission line between Hoover Dam and the city of LA had a higher voltage on the receiving terminal than it did at the Hoover Dam terminal.

A rectifier substation can be converted to accept regenerated power by replacing the the rectifiers with thyristor dual converters. This effectively involves replacing each diode with two thyristors in anti-parallel (one thyristor connected in the same direction as the original rectifier diode and the other connected in parallel but opposite direction of the first). The thyristors would have different firing circuits. Main reason for not doing this is that the power factor can be pretty low which makes the power company unhappy.

  • Member since
    September 2003
  • 16,282 posts
Posted by Overmod on Tuesday, November 3, 2020 7:39 AM

MidlandMike
AFAIK the NEC cat south of NY is still 25 Hz, so regeneration back into the 60 Hz grid would need that phase conversion.

If I remember correctly (it has been a long time) most - perhaps all by now - of the 25Hz power is frequency-converted rather than generated.  I don't remember whether those converters can be 'reversed' but my recollection is that they could.

An interesting modern point about regenerated AC is that the phase information is easily derived from the 'sink' power within a couple of cycles, and of course easily phase-shifted for any desired power-factor correction with the equivalent of simple diagnostic circuitry.  This was a fundamental part of the proposed German use of the enginion AG "zero emissions engine" in distributed standby generation.

My personal suspicion is that some combination of wayside power charging and onboard 'hybrid' storage a la Rail Propulsion Systems is a much better answer than regen all the way back to 'grid' 60Hz for polyphase transmission.  Of course the regen feeding 'consumer power' plays better in politics.  For DC 750V see the discussion of HESOP in the paper Peter Clark linked over on the Classic Trains forum:

https://www.engineersaustralia.org.au/sites/default/files/resource-files/2017-01/sydney_light_rail.pdf

  • Member since
    June 2011
  • 12 posts
Posted by casey56 on Wednesday, December 9, 2020 11:04 AM

Not quite related, but I read in a UK publication about London Underground that when feasible the track profile puts stations on top of "hills" so that gravity both slows incoming trains and accelerates departing trains.

Irv Smith, Missouri City TX

  • Member since
    March 2016
  • From: Burbank IL (near Clearing)
  • 12,382 posts
Posted by CSSHEGEWISCH on Wednesday, December 9, 2020 11:54 AM

casey56

Not quite related, but I read in a UK publication about London Underground that when feasible the track profile puts stations on top of "hills" so that gravity both slows incoming trains and accelerates departing trains.

Irv Smith, Missouri City TX

 
The CTA's South Side leg of the Green Line was built in the 1890's with a slight roller coaster profile for the same reason.
The daily commute is part of everyday life but I get two rides a day out of it. Paul
  • Member since
    December 2007
  • 1,260 posts
Posted by Falcon48 on Wednesday, December 9, 2020 5:01 PM

daveklepper

All PCC cars used dynamic braking, including CTA's "Spam Cans (6000s).  More modern equipment generally uses regerative braking or a combination of regen. and dynamic.  Often, as on all pre-WWII PCCs, this is in combination with air, and streetcars usually have magnetic track brakes for emergency stops.       Rapid transit cars, even PCCs, usually do not.

By PCC, I refer to trucks, motors, control system, not the PCC body style used on streetcars.

 

The Chicago PCC rapid transit cars had magnetic track brakes.  So do the modern Chicago cars. 

  • Member since
    September 2008
  • 1,094 posts
Posted by aegrotatio on Thursday, December 10, 2020 10:15 PM

The Washington Metro, a contemporary to BART, uses regenerative and dynamic braking, but whether regenerative braking is used to power nearby trains is disputed.  The original system called for wayside battery banks to store this energy but conventional wisdom suggests that they never really worked and dynamic braking is done using resistors in overhauled and newer cars.

I have not been able to find any further information about the *alleged* regenerative braking of the Washington Metro.

I can confirm that most, if not all, underground stations are located "uphill" of the ruling grade so that slowing down and speeding up to and from stations are assisted by gravity.

  • Member since
    January 2019
  • 951 posts
Posted by Erik_Mag on Friday, December 11, 2020 12:00 AM

At least back in the 1970's, BART used a blended regenerative/dynamic braking system, where resistors would be cut in as the third rail voltage increased above 1KV. With a brand new installation, I would be tempted to place ultracapacitor banks at stations to absorb the regenerated power and which would provide an assist with accelerating the trains after the stop.

  • Member since
    September 2003
  • 16,282 posts
Posted by Overmod on Friday, December 11, 2020 2:20 AM

aegrotatio
I have not been able to find any further information about the *alleged* regenerative braking of the Washington Metro.

Start here (PDF file).  Note that this system had the capacity to move 6-car trains, albeit at restricted speed (about 10mph).

I haven't been keeping up with the relative cost of supercaps and their associated packaging and equipment for wayside storage (a common acronym when searching is WESS) but I think the case for kinetic storage may still be better for transit applications where strong regenerative braking is relatively shortly followed by heavy acceleration draw.  Some of the ALPS technology like the use of a compulsator may be useful in these.  See the various descriptions of the Vycon WESS that was tested on the Los Angeles Metro system about eight years ago; if I recall correctly, this particular demonstration project captured about 2/3 the energy of regenerative braking as subsequent traction current.  These have little restriction posed by overheating, chemical effects, or spikes, and (when using magnetic bearings with appropriate guidance) have an essentially unlimited number of high-current cycles.

Rail Propulsion Systems [https://railpropulsion.com/] is working on a self-contained energy storage system as an 'add-on' to existing commuter locomotives.  Some of the stuff they're doing is very interesting...

  • Member since
    September 2008
  • 1,094 posts
Posted by aegrotatio on Friday, December 11, 2020 9:54 PM

Wow, that's very recent and very interesting, thanks.

The battery/capacitor energy storage systems I'm describing were allegedly installed in the 1970s and 1980s which I've found no evidence of ever happening.

 

  • Member since
    January 2019
  • 951 posts
Posted by Erik_Mag on Saturday, December 12, 2020 12:39 AM

I have a strong preference for ulttracaps over batteries for this application.

#1 reason is that an ultracap will happily float on any voltage that's less than the maximum continuous voltage, whereas most batteries want to float at a specific voltage which usually varies with battery temperature.

#2 ultracaps have very high energy efficiency, and that efficiency does not vary with terminal voltage.

#3 ultracaps can be fully discharged, which can be a significant safety issue. Lithium batteries can be a bit scary to work with as the short circuit current can be quite high - fully discharging a lithium battery basically destroys it.

#4 Ultracaps have a very high cycle count, I've seen 500,000 cycles bandied about.

Cost per watt hour of storage caoacity is pretty high, but the cost per watt hour with respect to storage capacity times cycle life is pretty low - running about $0.05/kw-hr for the 48V/165F modules selling for circa $1400 in 2015.

  • Member since
    September 2003
  • 16,282 posts
Posted by Overmod on Saturday, December 12, 2020 7:42 AM

Batteries in the normal sense are as poor for wayside service as they are for switch engines... and for the same reasons: high current inrush and draw requirements, cooling, and cumulative damage among them.

I remember going down to look at the first Metro operation with a group from the Princeton transportation program (I don't remember the year precisely, but mid-Seventies) and I think there were vaults designated as being for the peak recuperation storage.  In those days, energy storage in capacitors was very expensive per farad, and these might have been big electrolytics (which are really more like batteries with a different internal structure from a maintenance and cooling standpoint) -- the design idea, quite sensible in itself, being that the caps would 'charge-buffer' the peak heavy inrush current and then be discharged at an optimized rate into the cells of the battery, or charged on demand to allow high current to the train for starting.  This is enormously easier with super/ultra/hypercapacitors (the amusing prefix 'grade inflation' being seen in supercritical steam plants and the like too) but up to recently the devices themselves have relatively low maximum voltage (being nanofabricated with very small dielectric cross-sectional dimensions) and are only partially if at all self-healing if that is exceeded, so you needed enormous strings in massive parallel to get 'traction' current and voltage and spike (etc.) accommodation.

There was renewed interest in this around the turn of the century with magnetic storage.  Much of this at the time was utility-grid oriented but I had a technical CD sponsored by some government agency, I think in the late 1990s, which covered it as a mature technology with only minimum cryo requirement -- as presented this should have been a pervasive breakthrough and hasn't been, so likely someone was a bit overzealous, but one of the first things that came to my mind was vaults on the Metro...

There are some recent things like "liquid metal batteries" that could probably be used for WESS and I encourage anyone interested in this to Google around for alternatives.  As with the batteries used as "weights" for that weird gravity-energy-storage plan out West (where you had batteries the size of railroad cars being pushed uphill and turned sideways with off-peak power) these are usually crappy as actual batteries but so cheap (and in some cases so robust) that they make sense sitting where weight and size don't matter and cooling is much easier.

I continue to think that good magnetic-bearing KERS is a good competitive alternative for WESS as it should have similar extended cycle lifetime, and (I think) inherent spike resistance.  I tinkered a bit with integrating onboard KERS with gyroscopic 'negative cant deficiency' tilt on high-speed passenger trains when I was young and naive; there is much more sense in keeping it wayside.

  • Member since
    January 2019
  • 951 posts
Posted by Erik_Mag on Saturday, December 12, 2020 11:58 AM

I had thought that kinetic energy storage had some future, but the kicker was fatigue life. Did some doodling on the cycle life energy storage cost for the flywheel company in Mass, and got $0.50/kwhr. One caveat is that cycle life might be greatly extended if the units were typical "charged" to say 80% of capacity.

The awakening for me in regards to ultra caps (Maxwell's moniker), was a project I was working on in 2015. One power supply in the system needed to provide 10+kW bursts of power at 40 - 45VDC, but the customer requirements was no more than 13A of draw at 240VAC. Ran the numbers for the number of farads needed, the answer was greater than 30 - which suggested ultracaps. Figured that Digi-Key was worth a try for starters and found the Maxwell 48V/165F module - couldn't have found a much better solution...The icing on the cake for this project was that XP Power had a closed frame switching power supply with adjustable constant current limit as well as an adjustable voltage limit - no need for precharge and made it easy to meet the 13A line current draw limit.

Just for grins, I calculated cycle life energy storage costs of the Maxwell module to be $0.05/kw-hr, but inly for an application that was cycling at least 6 times per hour or more. Figure a switcher installation could pay for itself in a combination of fuel savings, better acceleration and reduction in brake wear.

Batteries aren't completely out of the question for wayside storage, but they would almost certainly require a regulator between the battery and the contact conductor (3rd rail or trolley wire). For a ~600VDC system, I would be very tempted to use 800V battery packs developed for EV's and a SiCFET half bridge with the battery connected to the upper rail of the half-bridge. In this case, the half bridge would act as a buck regulator when supplying power from the battery and a boost regulator when charging the battery from the traction power supply. Neat thing about this arrangement is that the control system would only need to make slight adjustments in timing to accomodate the change between buck and boost. Regulator efficiencies should be on the order of 99%. Each battery pack would get its own regulator to accomodate differences in battery voltage (as dictated by the battery management system) as well as making the system a lot more modular.

{N.B. Thinking about the battery voltage vs contact conductor, the battery voltage should be higher than the traction supply voltage. In the case of the battery voltage being normally lower than the traction voltage, a short circuit on the traction supply would lead to the battery discharging through the body diode of the high side FET with the likely distruction of the FET. In the case of the higher battery voltage, conduction through the body diode would only happen with a surge on the trolley/3rd rail.}

One other advantage of the EV battery wayside energy storage is that it would allow for orderly system shutdown (i.e. trains get to stations) in case of a blackout.

What would be peachy is if someone could come up with a workable sodium sulfur battery that used relatively common materials for the electrodes and separators.

  • Member since
    March 2021
  • 59 posts
Posted by Former Car Maintainer on Thursday, April 22, 2021 5:57 AM

Erik_Mag

At least back in the 1970's, BART used a blended regenerative/dynamic braking system, where resistors would be cut in as the third rail voltage increased above 1KV. With a brand new installation, I would be tempted to place ultracapacitor banks at stations to absorb the regenerated power and which would provide an assist with accelerating the trains after the stop.

 

BART used dynamic braking if the 3rd rail was not receptive and regenerative braking if the third rail was receptive. The idea of a train accelerating out of the station as another train was slowing to arrive was the idea behind "effective" regeneration. Unfortunately, due to headway constraints by the ATC system, regeneration wasn't very effective. The use of an onboard capacitive bank was used on the AC car versions to keep the intermediate link stiff through short rail gaps but never used to store regenerated energy during braking. The DC cars dynamic brakes were stepped in notches that were speed dependent, with loading decreasing as car slowed down. At about 10 mph the friction brakes took over. The dynamics on an AC car was controlled by switching an IGBT. It was either dissipated as heat , diverted to third rail or blended.Could capacitor banks be installed at the power stations? Dunno...the BPS system has a break even 10year period, about the lifespan of a Tesla car battery...I'm not convinced it would last 20...

Join our Community!

Our community is FREE to join. To participate you must either login or register for an account.

Search the Community

Newsletter Sign-Up

By signing up you may also receive occasional reader surveys and special offers from Trains magazine.Please view our privacy policy