25kV AC vs 750V DC

[Bald Rick]

And yet there is relatively little said here about the major problems on the North London line as well as via Tottenham Hale due to yesterday's OHLE supply issues. These stranded around 6 trains at the peak of the incident, I believe, including a 378 stuck between stations, on board which the air handling systems failed almost immediately, resulting in multiple passengers feeling unwell. So it goes to show that sometimes, despite trespassing causing more delays to 3rd rail electrified lines, it is impossible to avoid delays on electrified routes for many other reasons.

Had the supply systems for the NLL service had greater redundancy, I suppose that may well have been of assistance. From my personal experience, other than emergency isolations and blankets of ice, electrical supply infrastructure problems on 3rd rail routes generally seem to limit themselves to fairly small areas, meaning that routes may only be closed for a fairly short distance, once said hook switches are dealt with. Of course, if there are too few crossovers (a major bugbear of mine), this makes less of a difference, because starting & terminating short is not as readily made an option.

Unfortunately that was a relatively rare failure of the contact wire. No amount of redundancy would have helped. Only 1 train was stranded away from a platform.

 

[edwin_m]

Unfortunately that was a relatively rare failure of the contact wire. No amount of redundancy would have helped. Only 1 train was stranded away from a platform.

But was it due some problem with the design or maintenance of the wire or the pantograph? Rhetorical question. No answer required.

 

[W230]

Almost every DC compatible unit delivered here in the past decade or so has had the power artificially limited on the DC as it would overload the traction supply system. Compare a 377/5 at full acceleration on the Brighton main line vs the MML.

Interesting. When driving 377/5s (and /2s) the difference in AC and DC performance is noticeably large. Particularly when on a 12 car south of the river which seems very sluggish.

But I had always thought that this was simply down to less power being available on DC (due to lower capacities and it being dragged down by other units in the vicinity) but actually the 377s are artificially limited then to deal with the overcurrent issues that would present themselves? Well, I learn something new every day!

 

[notadriver]

Yes they are current limited when in a 8 car formation or longer on DC. A 7 car formation has similar spritely performance to a 4 car. Wasn't the DC third rail system originally designed for relatively lighter trains operating at a 75 mph maximum speed ?

 

[apk55]

When the low voltage DC system was originaly introduced trains were much slower, shorter and ligher than what we require today. A typical train on many lines would be 8 cars 1/4 axels motored and a peak accelerating current of around 1600A . Now we expect to run a much heavier 12 car train with higher acceleration taking around 3600A or more on some sections up to 7200A. Substations were often placed at the maximum distance apart which which ment upgrading is expensive and the only practical way has been to double the number.

 

[O L Leigh]

Swt and Dave - many thanks for taking the time to answer my question! X

Lou,

While the theory as explained by SWT and Dave is not incorrect, it doesn't take into account the way that return current is actually dealt with.

To answer your initial question, yes there is a return path but no it is not the track. An electric train will initially pass the return current out of it's wheels into the track but it passes very quickly into the return wire that often hangs on the outside of the OLE structure. The track is electrically bonded to this wire using "red bonds" (so called because they are painted red).

The reason why the return current doesn't hang around in the track at dangerous voltages is down to the booster transformers that you can also see mounted to the outside of OLE structures at various locations. These are connected to both the return wire and the live wire at 25kV AC. My understanding is that they are wound identically on both sides giving a ration of 1:1. When energised by a train entering the electrification section, the live side of the transformer will induce a current in the side connected to the return wire. This acts like a vacuum cleaner sucking up the return current out of the track and "boosting" it back along the return wire to the feeder station.

The other reason why there are not dangerous voltages in the track is because electric trains actually only require a small proportion of the 25kV AC, therefore the return voltage is significantly lower. As I have mentioned already, a Cl317 EMU at full bore only requires something like 600V DC.

It is perfectly safe to touch, work on or be around track under the OLE. There are no hazards to life posed by the electrical energy contained within it. The only dangerous part at ground level is a "red bond" if it has become detached, but that really is all.

As to the flooding question:

Yes flooding above the level of the railhead would be a problem anywhere. However, what it would not require is an isolation of the traction current over a much wider area. If the 3rd rail is in contact with the flood water there is a good chance that the flood water would also be electrically live (assuming that the traction current hadn't already tripped). Isolation of the current at the location of the flood would affect all other areas within the same electrification section.

O L Leigh

 

[snowball]

The other reason why there are not dangerous voltages in the track is because electric trains actually only require a small proportion of the 25kV AC, therefore the return voltage is significantly lower. As I have mentioned already, a Cl317 EMU at full bore only requires something like 600V DC.

I'm not an electrical engineer but that bit sounds wrong to me. An AC locomotive or EMU contains a step-down transformer to reduce the 25kV to the sort of voltage required by the traction motors. The overhead line is connected to one end of the primary winding and the track is connected to the other end of the primary winding. The fact that the voltage is lower on the secondary side is surely irrelevant to what the track sees.

 

[edwin_m]

I'm not an electrical engineer but that bit sounds wrong to me. An AC locomotive or EMU contains a step-down transformer to reduce the 25kV to the sort of voltage required by the traction motors. The overhead line is connected to one end of the primary winding and the track is connected to the other end of the primary winding. The fact that the voltage is lower on the secondary side is surely irrelevant to what the track sees.

Quite.

There could quite easily be dangerous voltages in the track at 25kV relative to nearby metal objects, even with no trains nearby. Hence all metal objects within touching distance are bonded to the rails so that the voltage between them and the rails (or connected things such as train bodies) will always be close to zero. However the booster transformers mean that very little return current flows through these bonds and the earth. and I believe the autotransformer system does the same in a slightly different way.

The same risk arises with DC electrification which is not intentionally bonded to earth for reasons of limiting stray current as explained by others above. These systems have special voltage limiting devices between the rails and metal objects, which normally don't conduct but would do so if the voltage exceeded a safe limit.

 

[Taunton]

I have a faint recollection that when the North London Line was electrified on third rail in the 1980s there were some obscure problems with stray currents affecting signalling down on the Victoria Line at Highbury, about 80 feet below, and these issues were behind the partial (and eventually complete) conversion of the line to 25Kv.

Can anyone explain what the issues were.

 

[starrymarkb]

The NLL has been 3rd rail through highbury for many years (possibly before the Victoria line)

The line was converted to dual AC (Freight) and DC (Passenger) lines in the 1980s, and then was converted to AC only from Hackney to Dalston and Camden to Acton (The latter was for North of London Eurostars and Freight)

Conversion to AC through Highbury only took place as part of the ELL works (and there is still DC in the ELL platforms)

 

[snowball]

Booster transformers are briefly described, with a circuit diagram, on pages 6-7 (7-8 as seen by the PDF) of this document.

It's quite old so it doesn't mention the 50kV autotransformer system.

 

[DownSouth]

I'm not an electrical engineer but that bit sounds wrong to me. An AC locomotive or EMU contains a step-down transformer to reduce the 25kV to the sort of voltage required by the traction motors. The overhead line is connected to one end of the primary winding and the track is connected to the other end of the primary winding. The fact that the voltage is lower on the secondary side is surely irrelevant to what the track sees.

You are correct and OL Leigh was wrong. The voltage going through the rails to the return wires is 25kV, but the current will be low - even an electric loco with power in the order of 5MW (~6800hp) will only draw a maximum current of about 200A

 

[Bald Rick]

Anyone who has never, or rarely, heard of problems with third rail may like to try and get a train between Wimbledon and Streatham now....

 

[Taunton]

The NLL has been 3rd rail through Highbury for many years (possibly before the Victoria line)

The line was converted to dual AC (Freight) and DC (Passenger) lines in the 1980s, and then was converted to AC only from Hackney to Dalston and Camden to Acton (The latter was for North of London Eurostars and Freight)

Conversion to AC through Highbury only took place as part of the ELL works (and there is still DC in the ELL platforms)

I am aware that the "old" NLL was converted from 4th to 3rd rail running some time in the 1960s, indeed about the time of the Victoria Line opening, but I understood that the stray currents issue arose from the extension from Dalston to Stratford in the 1980s - and they travelled quite some distance to interfere with the Underground. The interference was a considerable surprise to the project electrical engineers.

When the BART system first opened for trial running in San Francisco at the end of the 1960s, using 1000v dc third rail, stray currents were picked up by an unconnected university experiment into the conductivity of plants about 10 miles away from the line. It took some time to identify but the researchers knew they were man-made because they only happened on Mondays to Fridays.

 

[D7666]

Almost every DC compatible unit delivered here in the past decade or so has had the power artificially limited on the DC as it would overload the traction supply system.

Yes ... but I'm not sure that is a valid piece of evidence here ... is it ?

AIUI, heating, I-squared*R, or R I^2R is the limiting factor, not current overload.

DC traction units of all generations are always limited anyway. In conventional DC motor DC supply units with resistor you have a current limit relay. On a BR SR EE507 motor unit, the relay was set for 400 or 800 A to motor(s) depending on series or parallel motor connections.

All they have done on the modern AC motored units when on DC supply is use software to current limit the unit. But not for the same reasons.

DC motor units actually maximum overload supply more than AC motor units do i.e. when moving away from a standing start.

IIMU AC motor units are different in their current draw characteristic. In grossly over simplified terms , when all things are equal, a DC motor is a constant voltage machine while AC motors are constant current. In the latter, to get the constant current at the motor the electronics gubbins does the variable voltage variable frequency stuff and off course this draws a variable current from the traction supply, but this is a much flatter draw over the whole speed train profile than a DC motor unit.

The main reason for current limiting on the AC motor units on DC supply is really about I^2R heating in the DC lineside cables rather than actually overloading it. The peak load of a DC motor unit drops off very quickly so I^2R heat is not so much of an issue, the equipment has time to cool off before the next train, but an AC motor unit although I is nowhere near the same peak it is constantly at a higher value so obviously I^2 is high, and less time to cool between trains.


--
Nick

 

[Muzer]

Yes ... but I'm not sure that is a valid piece of evidence here ... is it ?

AIUI, heating, I-squared*R, or R I^2R is the limiting factor, not current overload.

DC traction units of all generations are always limited anyway. In conventional DC motor DC supply units with resistor you have a current limit relay. On a BR SR EE507 motor unit, the relay was set for 400 or 800 A to motor(s) depending on series or parallel motor connections.

All they have done on the modern AC motored units when on DC supply is use software to current limit the unit. But not for the same reasons.

DC motor units actually maximum overload supply more than AC motor units do i.e. when moving away from a standing start.

IIMU AC motor units are different in their current draw characteristic. In grossly over simplified terms , when all things are equal, a DC motor is a constant voltage machine while AC motors are constant current. In the latter, to get the constant current at the motor the electronics gubbins does the variable voltage variable frequency stuff and off course this draws a variable current from the traction supply, but this is a much flatter draw over the whole speed train profile than a DC motor unit.

The main reason for current limiting on the AC motor units on DC supply is really about I^2R heating in the DC lineside cables rather than actually overloading it. The peak load of a DC motor unit drops off very quickly so I^2R heat is not so much of an issue, the equipment has time to cool off before the next train, but an AC motor unit although I is nowhere near the same peak it is constantly at a higher value so obviously I^2 is high, and less time to cool between trains.


--
Nick

I would submit evidence that whenever they've introduced more power-hungry units recently (Desiros, Electrostars, etc. - presumably to do with the air conditioning) they've had to have a massive programme of substation upgrades and delay introduction of the new units until said upgrades are complete.

 

[D7666]

I would submit evidence that whenever they've introduced more power-hungry units recently (Desiros, Electrostars, etc. - presumably to do with the air conditioning) they've had to have a massive programme of substation upgrades and delay introduction of the new units until said upgrades are complete.

Yes

That power upgrade was exactly for this - nonetheless, there are still current limits on the EMUs.

The power reinforcement in simple terms generally allows AC motor 4500 Amp trains in the country areas and 6000 Amp trains in the suburban area, and capacity sized for the right number of train in each electrical section.

But you would never have been able to do any of that without the reinforcement. Yes the programme is complete, but the size of the equipment under that programme is very closely allied to current limiting on every EMU. The massive new DC substation at Edgware Road for the surface lines is similar evidence for S-stock, but the latter is still current limited.

--
Nick

 

[RichardN]

Wouldn't modern units benefit from a bit of on board battery to get them moving? Presumably this would help with finding a load for the brakes too.

 

[Pumbaa]

Anyone who has never, or rarely, heard of problems with third rail may like to try and get a train between Wimbledon and Streatham now....

Fnarr! Did raise a chuckle, having read this thread and seeing the system spit that one out.

 

[edwin_m]

Newer trains are heavier than older ones so have to use more power and therefore more current to maintain the same performance. Aircon also uses a significant amount of power. The AC network generally has few problems supplying this power (although extra feeder stations have been installed/proposed on some routes) but most parts of the DC system are near their limits.

A battery hybrid booster might help a bit on third rail routes, but it would make no difference on 25kV because that system is able to supply all the power the motors need. Almost certainly not enough benefit to justify the extra cost and weight of the battery.

Incidentally the 1.5kV DC system in the Netherlands is also close to its limit, with multiple overhead conductors to supply the current in some places. I believe they are also looking at 25kV conversion.

 

[O L Leigh]

I'm not an electrical engineer but that bit sounds wrong to me. An AC locomotive or EMU contains a step-down transformer to reduce the 25kV to the sort of voltage required by the traction motors. The overhead line is connected to one end of the primary winding and the track is connected to the other end of the primary winding. The fact that the voltage is lower on the secondary side is surely irrelevant to what the track sees.

Pardon me but, as you say, you are no electrical engineer.

There is no direct connection between the transformer primary winding and the track. The primary winding is connected only the OLE via the pantograph and the voltages are stepped down by the secondary and tertiary windings, rectified where necessary for DC systems such as the battery charger and traction packages, and then out through the wheels via the traction motor earth connections into the track at a much lower voltage.

Beside, you only have to see the difference between the earthing between the 25kV OLE and the return wire and track. The return wire is not always rubber sheathed except where it passes close over signalposts or other structures, but it's attached to the side of the OLE structure using just a single "pot" whereas the register arms have five "pots". The track is electrically insulated from the chairs only by a rubber mat and from the panrdol clips by a couple of plastic caps. How can these differences in insulation permit the same voltages?

There could quite easily be dangerous voltages in the track at 25kV relative to nearby metal objects, even with no trains nearby. Hence all metal objects within touching distance are bonded to the rails so that the voltage between them and the rails (or connected things such as train bodies) will always be close to zero. However the booster transformers mean that very little return current flows through these bonds and the earth.

Voltages induced in nearby metal objects by the magnetic field from the 25kV AC OLE are indeed a potential risk, which is why structures are bonded as you describe. It's also why OLE sections isolated to permit engineering work are also temporarily bonded to the nearest structures. However, these bonded structures feed into the same return path that is used for traction return.

While the induced voltage could be as high as several kilovolts, it certainly wouldn't be as high as 25kV. There certainly were issues with the dual-voltage 6.25kV AC/25kV AC OLE system installed in the early 1960s. When EMUs started blowing up their transformers it was discovered that, among other problems, there were potential issues with the voltage changeover switches on the units causing transient over-voltages on the transformer cores. When tests were carried out it was discovered that an isolated section of OLE would have an induced voltage of, strangely, around 6.25kV if the OLE on the neighbouring line was still energised at 25kV AC unless it was earthed.

The voltage going through the rails to the return wires is 25kV, but the current will be low - even an electric loco with power in the order of 5MW (~6800hp) will only draw a maximum current of about 200A

And by what mechanism is the voltage stepped back up again to 25kV AC?

O L Leigh

 

[swt_passenger]

There is no direct connection between the transformer primary winding and the track.

That has to be challenged. The transformer primary must have a connection to the track - the transformer won't work otherwise.

 

[O L Leigh]

That has to be challenged. The transformer primary must have a connection to the track - the transformer won't work otherwise.

Are you sure?

Don't forget that we're talking about alternating current not direct current. There does not need to be a potential difference in the way there would with a DC circuit with either a neutral connection (earth) or a connection to the other side of the power source, as the potential difference itself is provided by the alternating nature of the current. In order for the transformer to work all that is required is that the primary winding is connected to a source of AC power.

Also bear in mind that the OLE can induce a current in other electrical structures simply by being energised without an earth.

O L Leigh

 

[Tomnick]

I'm no electrical engineer either, but I'm with the others on this one - voltage is the potential difference between two parts of the circuit, so surely has to be equal at all points between the OLE and earth, or wherever it returns to, if that's the potential difference being considered. It doesn't differ from DC in that way, does it? If it did, you wouldn't need three pin plugs at home - one would suffice if you didn't need an earth!

 

[swt_passenger]

Are you sure?
...In order for the transformer to work all that is required is that the primary winding is connected to a source of AC power.

Yes I am 100% sure. The primary winding has to be connected to both the 25 kV supply and to a return path back to the 'shoreside' supply transformer (through the rails and autotransformer feeder system as applicable) or no current will flow through the primary winding...

It may not be shown with a direct connection in simplified circuit diagrams, but it is definitely there in the real thing.

 

[cjmillsnun]

Are you sure?

Don't forget that we're talking about alternating current not direct current. There does not need to be a potential difference in the way there would with a DC circuit with either a neutral connection (earth) or a connection to the other side of the power source, as the potential difference itself is provided by the alternating nature of the current. In order for the transformer to work all that is required is that the primary winding is connected to a source of AC power.

Also bear in mind that the OLE can induce a current in other electrical structures simply by being energised without an earth.

O L Leigh

I'm sure you're wrong on this. If you've ever seen overhead power lines on the National Grid, each branch of the transmission tower carries 4 cables. 3 of them are the phases (each of them AC), the fourth is the neutral which carries the return current.

There are normally six branches on each tower, with an earth at the top provided purely for safety (mainly as a lightning conductor)

To complete an electric circuit you must have a line conductor and a neutral return, doesn't matter whether that's AC or DC.

Therefore the others are correct in that the return must be 25kV and that is through the rails.

 

[DaveNewcastle]

Yes, the current path is a complete circuit, from the lineside supply transformer, then through either the overhead cable or 3rd rail, through the primary windings of the on-board transformer, through the chassis and wheel to the rails and then (and this is where we can get confused) a combination of multiple earthing bonds from the track to the return cable, and back to the lineside supply transformer.

The currents are equal in all parts of the circuit except where there are multiple paths, in which case the current will be lower in each path. There are multiple paths in the track and its bonding to the the return cable.

The voltages are normally considered to be relative to a zero at earth, and are highest on the overhead or 3rd rail conductor, drops through the on-board transformer, and reaches something close to zero in the rail and return cable.

The need to treat the rail and return cable with caution is for 2 reasons, Firstly, in normal operation there is a finite resistance in the return cable and therefore a finite voltage difference (not likely to be hazardous, even under high demand, but I have measured over 30 volts in return conductors relative to the earth to which they are ultimately connected); Secondly, in fault conditions, the full supply voltage may find an abnormal path to the return, such as a short in a transformer, a foreign object between the pick-up and chassis, in which case there is a brief high current and high voltage in the return circuit before the circuit breaker operates to isolate the supply. Due to the low probability of these fault conditions, and the brief duration of any hazardous voltages before protective devices operate, it is neither appropriate nor helpful to provide low leakage insulation between parts of the return circuit and ground.

Let's not confuse the current path of 3-phase distribution with single phase circuits!

. . . . overhead power lines on the National Grid, each branch of the transmission tower carries 4 cables. 3 of them are the phases (each of them AC), the fourth is the neutral which carries the return current.

If that were true, then I would imagine that the 'neutral' or return conductors would require three times the capacity of each of the three phase conductors. In fact, the earth conductor linking the top of each support pylon is a single conductor while the phases will have 1, 2, 4 or even 8 conductors each. The return current of any one phase flows largely through the loads and the phase supplies of the other two phases (hence the need to balance the demand across phases).

Also bear in mind that the OLE can induce a current in other electrical structures simply by being energised without an earth.

Indeed it can, but not normally a usefully high current. It may induce hazardously high voltages, but even a modest load (such as a suitably rated filament light bulb) will be enough to bring that induced voltage down to nothing - unless there is a complete circuit (such as a transformer connected between supply and zero).

 

[O L Leigh]

I'll be honest and say that I've been reading up on transformers as I post and have only just settled on the Railway Technical webpages. This does indeed show that there is a connection from the transformer primary winding and the axle bushes as a return path. Clearly there were one or two deficiencies in my traction training. My apologies for that.

However, I was very dubious about the claim that there must also be 25kV AC in the track because of this connection because clearly this would be incredibly dangerous. There certainly aren't any instructions or warnings about dangerously high voltages in the track and I think DaveNewcastle has given an explanation about why this is.

O L Leigh

 

[edwin_m]

I'll be honest and say that I've been reading up on transformers as I post and have only just settled on the Railway Technical webpages. This does indeed show that there is a connection from the transformer primary winding and the axle bushes as a return path. Clearly there were one or two deficiencies in my traction training. My apologies for that.

However, I was very dubious about the claim that there must also be 25kV AC in the track because of this connection because clearly this would be incredibly dangerous. There certainly aren't any instructions or warnings about dangerously high voltages in the track and I think DaveNewcastle has given an explanation about why this is.

O L Leigh

Dave and others have explained most of your concerns - respect for electrical engineers as well as train drivers! Just to add a couple more points.

Voltage is always measured between two conductors. There is no such thing as the voltage of a conductor on its own. This there is nominally 25000 volts between the 25kV OLE and the rails.

However because of the various bondings, all the metal structures near the track are electrically connected to it. These will therefore be at close to zero volts relative to the rails. Because many of these structures are also in electrical contact with the earth, the rails in turn are within a few volts of earth potential. Thus there is 25000 volts between the OLE and the rails, maybe 24990 volts between the OLE and the earth and only 10 volts between the rails and the earth.

As explained above, the return conductor (usually on the non-track side of the OLE poles) is connected to the booster transformers, which force the return current to travel back through it instead of through the rails and earth. This means that the voltage between the return conductor and the earth is also quite small, so the insulators don't have to face the full 25kV. They are there simply to stop the return current wandering off down the pole and back into the earth.

 

[DownSouth]

Wouldn't modern units benefit from a bit of on board battery to get them moving? Presumably this would help with finding a load for the brakes too.

My guess is that battery performance is not yet good enough for this to get out of the 'more trouble than they're worth' box without major reliability/safety issues.

It's not just the extra weight of the batteries themselves which would need to be considered, there would also be extra traction motors, extra control systems, extra fire suppression equipment, upgraded processing capabilities for the unit's brains, etc - all of which becomes deadweight the moment that the boost runs out.

You're more likely to see flywheel recovery systems on mainline trains than lithium ion batteries. A three tonne flywheel going to 8,000 rpm can store up to 33MJ of energy - enough to provide 33 seconds worth of 1,000kW boost, or 82.5 seconds worth of full power to a pair of 200kW traction motors - all without requiring any exotic minerals like batteries.

And by what mechanism is the voltage stepped back up again to 25kV AC?

O L Leigh

It doesn't need to go "back up" to 25kV AC because it's not a linear process of "conversions" - the basics of how a transformer works are covered in every half-decent secondary school physics curriculum, but evidently not in the training to become a train driver.

To put it simply, the the 25kV circuit passes through the train, in via the active lead and the pantograph and out via the wheels and the neutral lead going to the return circuit, with it doing work along the way - in the same way that a 240V circuit passes through a lightbulb from the active wire to the neutral wire in your house and does work along the way.

The work done by the 25kV circuit on the train is to induce a magnetic flux within the transformer's core. The transformer's core then induces a current within the secondary windings which have no electrical connection to the high voltage AC circuit.

The secondary windings for traction power are low voltage (therefore high current) circuits rectified to DC, which may be used to run DC motors directly or connected into an inverter to run AC motors. There will also be other secondary windings running off the same transformer - the number of loops in the wire dictate the voltage - for providing power to other onboard systems such as HVAC, lighting, control systems and so on.


As for it being dangerous to have low current 25kV running through the rails between the train's current location and the nearest connection from the rails to the return wires, others have explained all about bonding already, and it's a feature of AC electrification that it shuts itself down within a few milliseconds of the circuit getting broken. Your concern about inducing magnetic fields is irrelevant - we use much stronger permanent magnets to hold papers to our fridge doors!

However - it's probably useful to maintain a bit of mystique about the tracks being dangerous due to electricity because people still respect electricity thanks to seeing the effects but not the electricity itself - unlike trains which they may reason they can see coming. Maybe that's why your instructors kept you in the dark.