Polarity of 750Vdc electrification

Status
Not open for further replies.

HSTEd

Veteran Member
Joined
14 Jul 2011
Messages
13,137
So this occured to be a while back:

Most traditional DC loads found on trains would consist of resistive heating/lighting elements or series wound motors, either in motor generators or in traction systems. These loads are totally insensitive as to the polarity of the voltage applied to them and will function identically in either case as long as the magnitude of the voltage is correct.
The rest would mostly be shunt wound motors with the possibility of controlling shunt windings seperately using contactors.

So is there any particular reason to maintain a single polarity for the DC power voltage relative to the ground?
This would be especially useful on tramways and the like that use pantographs/trolley poles because you could have the up line at +750V and the down line at -750V, so that multiple vehicles on opposite lines will 'share' rail return current and hold down current leakage. Substation spacing could be improved significantly thanks to lower currents and lower earth voltage rise as the worst case would only have half the current flowing in the rails (max capacity on one line and no demand on the other).

This becomes more problematic on third rail systems because of the long train busses which prevent simple "phase brakes" as the two lines will short through the train bus.
But you could probably work something out involving coasting sections longer than the longest single unit (4 cars was the longest commonly used on SR I believe?) that would only be brought in in specific circumstances to assist stalled trains. For crossovers the coasting sections would obviously only be de-energised if the crossover was set to connect the two lines.
 
Last edited:
Sponsor Post - registered members do not see these adverts; click here to register, or click here to log in
R

RailUK Forums

HSTEd

Veteran Member
Joined
14 Jul 2011
Messages
13,137
Well yes, but I was more interested in the historical reason why this has not been done, as it would seem to be a reasonable way to significantly improve the electrical properties of the network without violating the traditional limits on tramway voltages.


With modern systems you could probably build a static converter that would functionally identically on both polarities - indeed if it was for auxiliary loads and did not need to regenerate then a full bridge rectifier would be sufficient.
 

Nym

Established Member
Joined
2 Mar 2007
Messages
8,573
Location
Somewhere, not in London
It would be indeed, but it introduces potentially significant losses into the system for no real reason. Around 1 - 1.5% losses for no gain I'd think.

Perhaps the reason was to reduce the equipment count or provide a level of redundancy, this being in the days of RCs and MARs. Perhaps this was so that a reduced number of larger RCs could be provided.

There is also the issue of if you draw different currents on each phase of a 'centre tapped' system, it can push the neutral point around in ways you wouldn't like, or possibly expect.
 

Bald Rick

Veteran Member
Joined
28 Sep 2010
Messages
17,964
Presumably to keep everything simple in complex areas / S&C. If the lines had different polarity, a train swapping from one to the other would experience 1500v across the gap (I think!).
 

HSTEd

Veteran Member
Joined
14 Jul 2011
Messages
13,137
Indeed, you would likely have to have complex areas all at one polarity and provide neutral rail sections on the appropriate entrance tracks.
Something like an ~85m neutral section with a pair of trackside knife switches so it can be brought to a desired polarity to help move a train out of it as required.

But it still seems like it would be able to significantly improve substation spacings.

EDIT:

With modern systems you could provide pairs of appropriate configured interconnected buck converters at intermediate points in a track feeding system that would rebalance current between the conductors such that no current flows in the return rails.
One converter would be configured with a 'reference earth' of -750Vdc, and would be configured to convert 1500V (relative) to 0V, discharging through an appropriate diode into the rail.
The other would be configured as a 'negative buck' with a 'reference earth' of +750Vdc, being configured to convert -1500V to -750V, also discharging to the running rail through an appropriate diode.

In other words a 1500Vdc auto transformer system. Buck converters are quite cheap really and it would not have any circuit breakers as it would use a tramway feeding topology, only make only contactors would be required.
 
Last edited:

HSTEd

Veteran Member
Joined
14 Jul 2011
Messages
13,137
Given how frequent the sub stations are, why not just rectify it to where you want it?

I am more thinking in the context of the "low cost electrification" study down by DeltaRail a few years back.

1500Vdc between conductors with 750V betwen conductors and nominal ground is still classified as low voltage by the IEC (indeed it is nearly the limit of such systems).

This combined with the idea of low power current limited 400V fed substations could potentially allow for relatively long 750Vdc routes without any high voltage equipment at all as it would allow power to be more efficiently transmitted "lengthways" down the alignment.
 

edwin_m

Veteran Member
Joined
21 Apr 2013
Messages
20,942
Location
Nottingham
The historic reason for making the running rails negative was that it reduced electrolytic corrosion due to stray current. I suspect traction electronics designed for DC wouldn't be able to cope with reverse polarity. A bridge rectifier could presumably be added to the inputs but would reduce efficiency a bit.

Also bear in mind that double track tramways have the two OLE wires bonded together and the losses in the wires would roughly double if this wasn't done.
 

HSTEd

Veteran Member
Joined
14 Jul 2011
Messages
13,137
Also bear in mind that double track tramways have the two OLE wires bonded together and the losses in the wires would roughly double if this wasn't done.

Losses from a single tram operating load would double however if we assume trams are common and equally distributed across both lines we go from a single 1000A (as an example) load with a 1 ohm OLE feeder to a pair of 500A loads with 2 ohm feeders.

Losses would be 1000W in the first case and 2x 500W in the second case, so the losses will tend to balance out.

After all there is half the current through double the resistance twice, which all balances out thanks to power lost being double the resistance. This makes sense since you can model a DC conductor as an array of thinner DC conductors quite reasonably.

Double track tramways that are relatively heavily used and have multiple vehicles in each feeder section would also see a reduction in rail voltage which would tend to reduce losses which can be surprisingly high in the track.

EDIT:
I just realised (whilst half asleep!) that my idea of interconnected buck converters is highly problematic because you could generate giant loop currents in them.
What you actually need is a reversible isolated DC-DC converter. There are several suitable topologies such as a DC-DC flyback that can use a compact high frequency transformer for relatively low cost and a compact form factor.

One side would be connected between zero volts and the +750V rail and the other would be connected between zero volts and the -750V rail.
The controll system would be programmed so that it always transferred power from the higher voltage side to the lower voltage side. Since the two power supply lines will have comparable resistances and thus voltage drops per unit current this will tend to force current out of the running rail and into the opposite polarity power cable.
 
Last edited:

Joseph_Locke

Established Member
Joined
14 Apr 2012
Messages
1,878
Location
Within earshot of trains passing the one and half
The historic reason for making the running rails negative was that it reduced electrolytic corrosion due to stray current. I suspect traction electronics designed for DC wouldn't be able to cope with reverse polarity. A bridge rectifier could presumably be added to the inputs but would reduce efficiency a bit.

Also bear in mind that double track tramways have the two OLE wires bonded together and the losses in the wires would roughly double if this wasn't done.

The fourth rail on LUL is actually at about -50V IIRC?
 

edwin_m

Veteran Member
Joined
21 Apr 2013
Messages
20,942
Location
Nottingham
The fourth rail on LUL is actually at about -50V IIRC?

The four-rail system is a special case where the running rails don't have a hard connection to the traction power but they do float at a voltage between that of the third and fourth rails. The fourth rail is therefore indeed at a negative voltage but I think it may be a bit more negative than -50V.
 

AM9

Established Member
Joined
13 May 2014
Messages
9,555
Location
St Albans
The four-rail system is a special case where the running rails don't have a hard connection to the traction power but they do float at a voltage between that of the third and fourth rails. The fourth rail is therefore indeed at a negative voltage but I think it may be a bit more negative than -50V.

Not quite. The LU DC supply is nominal 630V with the centre rail at -210V and the outer rail at +430V wrt the running rails. This split is optimised to reduce the effects of elecrolytic corrosion on steel/iron tunnel infrastructure and other subterranean conductive materials. The actual difference between the grounded running rails and the two conductors can vary slightly according to traction loads on the line and various leakages. The position of the tunnel/running rails is determined at the feed points as it has to be bonded to prevent a potential between train structures and the surrounding grounded infrastructure (think platforms).
Where the lines run onto NR metals, e.g. west of Queens Park or at Gunnersbury station junction, there is an isolated section and the centre rail is bonded to the running rails leaving the 3rd (outer for LU stock) rail powered with a nominal 630VDC, (typically at the inner London level of 660VDC).
 
Last edited:

MarkyT

Established Member
Joined
20 May 2012
Messages
5,037
Location
Torbay
Not quite. The LU DC supply is nominal 630V with the centre rail at -210V and the outer rail at +430V wrt the running rails. This split is optimised to reduce the effects of elecrolytic corrosion on steel/iron tunnel infrastructure and other subterranean conductive materials. The actual difference between the grounded running rails and the two conductors can vary slightly according to traction loads on the line and various leakages. The position of the tunnel/running rails is determined at the feed points as it has to be bonded to prevent a potential between train structures and the surrounding grounded infrastructure (think platforms).

The running rails cannot both be grounded, as on all but the latest axle counter equipped Underground lines continuous track circuits are fitted which need the rails to be insulated from each other. It is possible that one rail may be grounded and the other could be insulated.
 

Taunton

Established Member
Joined
1 Aug 2013
Messages
5,999
I've never understood (and would welcome the input of an electrical engineer) how substantial traction currents can "return to earth" through the running rails, without having any impact on the track circuits, which seem to operate at a notably low voltage over long distances without apparently significant voltage drop, and yet are sensitive enough and reliable enough to operate critical signal equipment without being affected by the return currents in the same rails.
 

hwl

Established Member
Joined
5 Feb 2012
Messages
6,459
I've never understood (and would welcome the input of an electrical engineer) how substantial traction currents can "return to earth" through the running rails, without having any impact on the track circuits, which seem to operate at a notably low voltage over long distances without apparently significant voltage drop, and yet are sensitive enough and reliable enough to operate critical signal equipment without being affected by the return currents in the same rails.

The track circuits in 3rd rail land were traditionally AC with large impedance bonds (which allow DC but not AC to pass through) and insulated block joints between the track circuit areas. The currently would usually only have to return via the running rails for <1.5miles to the nearest substation or TP hut i.e. across several track circuit areas.

The impedance joins have a habit of blowing up spectacularly (see New Cross failure a fortnight ago that stopped all services to/from Charing Cross and melted the plastic in the block joint). Impedance bond failures cause a lot of "signal failures" is 3rd rail land.

The modern track circuits use higher frequency AC (aka Audio frequency) allowing each TC in the UK to have its own frequency eliminating the need for most block joints and impedance bonds.
 

MarkyT

Established Member
Joined
20 May 2012
Messages
5,037
Location
Torbay
The track circuits in 3rd rail land were traditionally AC with large impedance bonds (which allow DC but not AC to pass through) and insulated block joints between the track circuit areas. The currently would usually only have to return via the running rails for <1.5miles to the nearest substation or TP hut i.e. across several track circuit areas.

The impedance joins have a habit of blowing up spectacularly (see New Cross failure a fortnight ago that stopped all services to/from Charing Cross and melted the plastic in the block joint). Impedance bond failures cause a lot of "signal failures" is 3rd rail land.

The modern track circuits use higher frequency AC (aka Audio frequency) allowing each TC in the UK to have its own frequency eliminating the need for most block joints and impedance bonds.

Even with no insulated block joints, the tuned 'jointless' type track circuits still need impedence bonds to allow return of the traction current to the substation without shorting out the rails. The only way to remove the electrical commonality between traction return and train detection is to substitute axle counters. Then you can cross bond and common up all the rails as much as you like to reduce the return current resistance. I suspect that LUs continued use of 4 rail power systems was as much to avoid the track circuit electrical commonality as to avoid the galvanic erosion leakage problem in the iron tunnels. Exploding signalling equipment in deep tunnels is clearly highly undesirable!
 

AM9

Established Member
Joined
13 May 2014
Messages
9,555
Location
St Albans
430 - (-210) = 640.

Maybe the 430 should be 420?

Ooops, should be 420 + 210. 630 is the LU nominal.
--- old post above --- --- new post below ---
Even with no insulated block joints, the tuned 'jointless' type track circuits still need impedence bonds to allow return of the traction current to the substation without shorting out the rails. The only way to remove the electrical commonality between traction return and train detection is to substitute axle counters. Then you can cross bond and common up all the rails as much as you like to reduce the return current resistance. I suspect that LUs continued use of 4 rail power systems was as much to avoid the track circuit electrical commonality as to avoid the galvanic erosion leakage problem in the iron tunnels. Exploding signalling equipment in deep tunnels is clearly highly undesirable!

That's assuming that the track is DC circuited. It is normal to use ac with DC inductive commoning which allows ac to be available for train detection.
 

Taunton

Established Member
Joined
1 Aug 2013
Messages
5,999
Thank you "hwl". If you use AC track circuits on DC lines (and presumably vice-versa), I wonder how does it work on tracks with both, such as the approach to Euston?

There was some substantial problem on the North London Line in the 1980s, not sure whether it was the conversion from 4th rail (like the Underground) to 3rd rail, or the extension that was also done at the time from Dalston to Stratford, but it had an impact on the Victoria Line signalling quite some distance away (and to the surprise of all concerned), and led eventually to the new 3rd rail being abandoned and dual voltage trains being drafted in to use the overhead wires which had been initially installed for freight services. Anyone recall what the issue was?

I do remember hearing that a tracked excavator going over the old main line level crossing at Silk Mill, west of Taunton, shorted out the track circuits between adjacent tracks.
 

snowball

Established Member
Joined
4 Mar 2013
Messages
4,950
Location
Leeds
Thank you "hwl". If you use AC track circuits on DC lines (and presumably vice-versa), I wonder how does it work on tracks with both, such as the approach to Euston?

There are more ways a track circuit can be immune to AC traction current than for it to be a DC track circuit. An AC track circuit can be immune to AC traction current provided the frequencies used are carefully chosen to avoid the traction frequency (50Hz) and harmonics of it.

--- old post above --- --- new post below ---

430 - (-210) = 640.

Maybe the 430 should be 420?

Who cares, third rail is for Scalextric, not for proper trains.

The question was (obviously) about the 3rd+4th rail system. Are you advocating conversion of LU to overhead wires? You'd have to make the trains even less tall than they already are.
 
Last edited:

HSTEd

Veteran Member
Joined
14 Jul 2011
Messages
13,137
LU could always switch to +/- 750V...... that would cause all sorts of fun.
 

dgl

Established Member
Joined
5 Oct 2014
Messages
1,444
Ignore original post, yes 750+750 would be fun, would please the enthusiast's though, as they would do a very good impression of either a steam engine or an old diesel :) .

Or as photon would say "let's put these trains on the big boys supply and see if we can turn the power up to make these trains run faster, oh no I've popped it"
 
Last edited:

Philip Phlopp

Established Member
Joined
31 May 2015
Messages
3,004
The question was (obviously) about the 3rd+4th rail system. Are you advocating conversion of LU to overhead wires? You'd have to make the trains even less tall than they already are.

If only it were possible...

Genuinely, as we all know, it would solve a rather irritating issue with the Metropolitan line which is likely to cause some trouble in the next decade.

It would have saved a bit of bother if the S-Stock for the line had a 25kV dual voltage capability from the outset, though I'm led to believe it could be provided if absolutely necessary.
 

snowball

Established Member
Joined
4 Mar 2013
Messages
4,950
Location
Leeds
Ignore original post, yes 750+750 would be fun, would please the enthusiast's though, as they would do a very good impression of either a steam engine or an old diesel :) .

Which post are you asking us to ignore? You haven't posted previously in this thread.
 

MarkyT

Established Member
Joined
20 May 2012
Messages
5,037
Location
Torbay
Thank you "hwl". If you use AC track circuits on DC lines (and presumably vice-versa), I wonder how does it work on tracks with both, such as the approach to Euston?

There are more ways a track circuit can be immune to AC traction current than for it to be a DC track circuit. An AC track circuit can be immune to AC traction current provided the frequencies used are carefully chosen to avoid the traction frequency (50Hz) and harmonics of it.

Yes there are some strange frequency AC track circuits used for mixed electrification areas. I remember seeing large numbers of 83.3Hz vane relay track circuits in the old (pre 1990s) signalling on the GE main line. These had been provided initially under the 1500VDC electrification but were suitable for retention when the line was converted to 25kVAC. They were all swept away when the current SSI signalling was introduced.

Here's a quote from a thesis found here:
http://www.iea.lth.se/publications/Theses/LTH-IEA-1021b.pdf

Initially, AC supplied track circuits were developed for use in DC supplied lines. They were fed from the public mains 50 Hz. However, with another frequency, for instance 83.3 Hz, AC track circuits can also be used on lines supplied with 50 Hz or 16 2/3 Hz. 83.3 Hz track circuits can also be used on DC supplied tracks if the DC supply contains too much 50 Hz harmonics, due to rectifier components asymmetries or if there is a high amount of 50 Hz vagabonding current.

Returning to Euston, the entire station and it's approaches were equipped with axle counters in the 1990s resignalling so track circuit compatibility with the two electrification systems is not an issue.

There was some substantial problem on the North London Line in the 1980s, not sure whether it was the conversion from 4th rail (like the Underground) to 3rd rail, or the extension that was also done at the time from Dalston to Stratford, but it had an impact on the Victoria Line signalling quite some distance away (and to the surprise of all concerned), and led eventually to the new 3rd rail being abandoned and dual voltage trains being drafted in to use the overhead wires which had been initially installed for freight services. Anyone recall what the issue was?

They were trying to run a significant area of mixed eletrification which is a particularly undesirable situation due to the conflicting earthing requirements of the two systems. With DC only electrification, the rails 'float' on insulated
pads, unconnected to earth or any other metalwork. This encourages the high DC traction current to return through the rails alone and not through any alternative paths through other metalwork and the ground itself. In turn this prevents some of the worst excesses of DC leakage current and the corrosion that may occur as a result. High voltage AC by contrast incorporates a lot of safety earthing including at least one of the rails (depending on the train detection system used) bonded periodically to all metal structures within a certain distance of the wires. This is a safety feature to ensure touch potential can never rise to levels dangerous to human beings: passengers and others at stations and crossings; maintenance workers and train crew out on the line. In a dual electrified area the high voltage concern has to take precedence and all the safety earthing has to be provided. This then gives the DC return current many many possible alternative paths away from the rails and effectively encourages the highly undesirable leakage and possible consequent corrosion. Apparently this was experienced over a wide area when the NLL dual electrification was attempted. Best practice where dual electrification is unavoidable such as at system changeover locations is to firstly keep the area as small as possible, and then use massive additional strengthening conductors to lower the desired return path resistance and encourage as much of DC current as possible to flow that way. Leakage can never be eliminated entirely though.

Note AC return current leakage via the multiple alternative paths of a comprehensively earthed system is not a major problem. For a given traction load, the current is much smaller due to the high voltage and AC current does not cause corrosion like DC anyway.
--- old post above --- --- new post below ---
I'm rarely on Met turf.
Which issue?

I think he may be talking about what happens when the diesel units on the Aylesbury service need to be renewed and what form of electrification might be used if (as seems likely) electric replacements are proposed. It is complicated with parallel sections currently unelectrified and other 4-rail sections shared by LUL and NR trains. If 25kV was chosen for north of Amersham, and from Harrow going south into Marylebone, then one solution could be new dual system electrics, changing over from OHLE to 4-rail at Amersham, then back to OHLE at Harrow.
 
Last edited:

talltim

Established Member
Joined
17 Jan 2010
Messages
2,386
For trams, surely the extra insulators required to separate the contact wires for each direction would give more hassle than any savings you get?
 

HSTEd

Veteran Member
Joined
14 Jul 2011
Messages
13,137
If only it were possible...

Genuinely, as we all know, it would solve a rather irritating issue with the Metropolitan line which is likely to cause some trouble in the next decade.

It would have saved a bit of bother if the S-Stock for the line had a 25kV dual voltage capability from the outset, though I'm led to believe it could be provided if absolutely necessary.
Considering S-Stock likely complies with LU regulations and thus has no 750v bus aboard the train, with each power module obtaining its power from its own pair of shoes, that sounds very expensive.

1500Vdc four rail would probably come out quite competitive with 25kV thanks to the huge reduction in currents compared to now, and it would have none of the normal earth return issues.

Really the best solution I can come up with is likely diverting everything possible over the CML and just electrifying the Amersham/Aylesbury line entirely at 4th rail and transferring it to the LU.
They would probably need a couple of platforms at Marylebone for the extra terminators though.

For trams, surely the extra insulators required to separate the contact wires for each direction would give more hassle than any savings you get?

That depends, many tram electrification schemes don't have insulators any more, they use parafil ropes to suspend the cables and can thus get away without them.
On a heavily used route it effectively turns the system into a 1500Vdc one with a big earth return conductor, which would likely significantly increase electrical stability in the system.
 
Last edited:
Status
Not open for further replies.

Top