Name for section of DC overhead electrification where power from two substations overlaps

[eldomtom2]

I am looking for the English name for this. Here is a diagram of what I'm talking about in case it isn't clear:

It's a Japanese diagram, so it uses the seemingly Japan-exclusive term "air section".

 

[Snow1964]

The diagram is where two wire lengths end with an overlap.

Where each is supplied by a separate substation will be a short neutral section, there is a short length of non-conductor linking so pantograph just runs across, but can't be in contact with both at same time. The circuit breakers are shut through these as a trackside balise tells train to turn off power, then turn on again other side.

There is a third type a short section at switches and crossovers so that power on one can be isolated, more an isolating switch. These are usually U shaped with other section running down middle but not touching

 

[59CosG95]

The diagram is where two wire lengths end with an overlap.

Where each is supplied by a separate substation will be a short neutral section, there is a short length of non-conductor linking so pantograph just runs across, but can't be in contact with both at same time. The circuit breakers are shut through these as a trackside balise tells train to turn off power, then turn on again other side.

There is a third type a short section at switches and crossovers so that power on one can be isolated, more an isolating switch. These are usually U shaped with other section running down middle but not touching

Sounds like the DC equivalent of a Carrier Wire Neutral Section (CWNS) to me.

 

[swt_passenger]

However DC systems are not necessarily always powered quite the same way as AC. On the DC Tyne Wear metro, long sections are fed in parallel from both ends, fed from two separate substations. Eg Monument to South Gosforth is “double-end fed”, then South Gosforth to Benton Square is also double-end fed.

So at the catenary overlaps between those points the wires either side ARE electrically connected. I don’t know if there’s a specific name for it though. Here’s a screen shot from a YouTube cab ride video:



This can have resulting problems with fault detection, there’s a RAIB report (that deals with a fire on a Metro unit at South Gosforth), that covers the power supply setup and the necessary fault monitoring in some detail:

https://assets.publishing.service.gov.uk/media/547c8fbded915d4c0d00015b/R182013_130925_South_Gosforth.pdf

 

[edwin_m]

However DC systems are not necessarily always powered quite the same way as AC. On the DC Tyne Wear metro, long sections are fed in parallel from both ends, fed from two separate substations. Eg Monument to South Gosforth is “double-end fed”, then South Gosforth to Benton Square is also double-end fed.

So at the catenary overlaps between those points the wires either side ARE electrically connected. I don’t know if there’s a specific name for it though. Here’s a screen shot from a YouTube cab ride video:

View attachment 142530

This can have resulting problems with fault detection, there’s a RAIB report (that deals with a fire on a Metro unit at South Gosforth), that covers the power supply setup and the necessary fault monitoring in some detail:

https://assets.publishing.service.gov.uk/media/547c8fbded915d4c0d00015b/R182013_130925_South_Gosforth.pdf

As far as I know, DC systems always put all the substations in parallel in this way. It means the next substation along can provide some of the current if two or three trains happen to be accelerating in the same area.

 

[swt_passenger]

As far as I know, DC systems always put all the substations in parallel in this way. It means the next substation along can provide some of the current if two or three trains happen to be accelerating in the same area.

I assumed it might be fairly standard, but hadn’t researched it online. Stands to reason though, as we already know the third rail DC operates in parallel over large areas for the same reason. I found that the RAIB report was a fairly straightforward explanation in any event, and I think I’ve quoted it before when a discussion of TW Metro power supply occurred.

 

[lukeychops]

Looks like a single span overlap

 

[ac6000cw]

As far as I know, DC systems always put all the substations in parallel in this way. It means the next substation along can provide some of the current if two or three trains happen to be accelerating in the same area.

Which I assume also means the distance between the substations can be increased (compared to if there was a neutral section at the midpoint, meaning no load sharing between adjacent substations)?

The Japanese diagram seems to imply the A & B catenary are not connected together, so the pantograph will be shorting them together briefly as it passes through the overlap?

 

[edwin_m]

Which I assume also means the distance between the substations can be increased (compared to if there was a neutral section at the midpoint, meaning no load sharing between adjacent substations)?

The Japanese diagram seems to imply the A & B catenary are not connected together, so the pantograph will be shorting them together briefly as it passes through the overlap?

I expect so. It also improves the benefit of regenerative braking, because it makes it more likely there is another train able to use the regenerated power. This won't work over an infinite distance due to voltage drops, but it works even less far if the line is sectioned electrically.

If the DC sections have to be electrically separated, then I'd have thought they would need to be separated absolutely by an insulated section, without the possibility of then being briefly electrically connected via the pantograph. During that time the electrical characteristics of the system would be significantly different, making it more difficult for the system to detect and respond to a short circuit.

@eldomtom2, do you have some context for the original diagram. Is it perhaps showing how not to do it?

 

[eldomtom2]

The Japanese diagram seems to imply the A & B catenary are not connected together, so the pantograph will be shorting them together briefly as it passes through the overlap?

Maybe? I'm not the most knowledgable on electricity. Here's a google translated Japanese description of an "air section":

Refers to the overlapping section where power flows from different substations and is DC electrified. In the case of the air section, each of the two overhead lines will be powered by a separate substation. Different substations do not supply exactly the same electricity, so the voltage is also different. Electricity has the property of flowing from a place with a higher voltage to a place with a lower voltage, so if there is a difference in voltage, electricity can flow in the air. This is called "arc discharge", and when arc discharge occurs, a large amount of electricity flows along with strong light. When a large amount of electricity flows, heat is generated, and this heat causes the overhead wire to break. There is no problem in "passing" the air section, but if the train "stops" in the air section, the heat from the arc discharge accumulates and cuts the overhead wire. In order to prevent such accidents, there is an "air section sign" to inform that the vehicle must not stop within the basic section, and in the unlikely event that the vehicle stops, it is necessary to take appropriate measures such as lowering the pantograph.

 

[edwin_m]

Maybe? I'm not the most knowledgable on electricity. Here's a google translated Japanese description of an "air section":

That quote confirms our understanding in previous posts is correct. I really can't see any reason to do it this way - the need to avoid stopping in the air section even removes the only possible benefit I can see, that it would be impossible for the train to be "gapped" with the pantograph in a neutral section. Is there any text nearby that explains why they might want to use this arrangement?

 

[ac6000cw]

I really can't see any reason to do it this way - the need to avoid stopping in the air section even removes the only possible benefit I can see, that it would be impossible for the train to be "gapped" with the pantograph in a neutral section.

I agree - at face value, it seems a silly thing to do.

A train very occasionally getting stranded in a neutral section is a minor problem compared to potentially melting through the conductor wire and possibly having a broken 'live' end touch the roof of the train...

 

[eldomtom2]

That quote confirms our understanding in previous posts is correct. I really can't see any reason to do it this way - the need to avoid stopping in the air section even removes the only possible benefit I can see, that it would be impossible for the train to be "gapped" with the pantograph in a neutral section. Is there any text nearby that explains why they might want to use this arrangement?

Here's a section from the Google translated Wikipedia article. From this it appears that trains can still draw power in "air sections":

An air section is basically a piece of equipment that is installed in a DC electrified section at the border between the section where the voltage is being fed from a substation and the section where the same voltage is being sent from another substation. At the ends of both overhead wires, insulators are installed in series near the retention points of both overhead wires, and the overhead wires are placed in parallel between the 1st and 2nd spans of the overhead wires using air as an insulator.

As a result, unlike a neutral section, even though the two overhead wires are nominally at the same voltage, there is actually a slight potential difference depending on the conditions. When an electric train passes through this area, the two parallel catenary lines are short-circuited by the
pantograph, but it is possible to pass at normal speed, power running, no problems with regenerative braking.

[...]

Electric trains that collect electricity from overhead lines must basically not stop within the air section. However, in the case of an emergency stop due to the operation of the level crossing warning device or the reception of the protective radio, the train may be forced to stop in the air section.

If the train stops in the air section, once lower all the pantographs in the formation. Then, when starting up, the pantographs in the formation are raised all at once and quickly pass through the air section (because there is no problem in raising the pantographs directly under the air section as long as it is within the allowable time).

 

[ac6000cw]

Here's a section from the Google translated Wikipedia article.

Please post a link to the article so we can see the whole context.

 

[eldomtom2]

Please post a link to the article so we can see the whole context.

エアセクション - Wikipedia

ja-m-wikipedia-org.translate.goog

 

[ac6000cw]

エアセクション - Wikipedia

ja-m-wikipedia-org.translate.goog

Thanks.

It's interesting that the article contains a list of wire melting incidents when trains have stopped in an 'air section', including a photo caption where it says "There are also scorch marks on the vehicle on the right caused by cut and hanging overhead wires."

 

[stuving]

Reading those descriptions, I can only make sense of them if the voltage difference between sections is quite large. Too large to allow parallel feeding, but OK for a short overload current while a train passes the air section. In that case this is not a bad simple solution, provided the procedural protection against shorting the air section for too long (telling drivers not to stop there) is acceptable.

Why the difference would be so large I have no idea - that's an explanation of the explanation, and probably to be found somewhere else.

 

[ac6000cw]

A carbon arc lamp needs less than 100V between the ends of the carbon rods (usually a few mm apart) to maintain the arc, after it has been started (struck) by touching the rods together. Arc welding uses similar or lower voltages to melt metals together.

So I don't think it would take much of a voltage difference between the sections to get a destructive arc going (and in a carbon arc it's over 3000C and copper melts at about 1100C).

(Off-topic, but I do find it amusing that modern movies still often use a 'schuu/sizzle' sound effect of an old/obsolete carbon arc lamp starting up whenever large spotlights are turned on, after the loud 'thunk' of a big switch of course...)

 

[swt_passenger]

Reading those descriptions, I can only make sense of them if the voltage difference between sections is quite large. Too large to allow parallel feeding, but OK for a short overload current while a train passes the air section. In that case this is not a bad simple solution, provided the procedural protection against shorting the air section for too long (telling drivers not to stop there) is acceptable.

Why the difference would be so large I have no idea - that's an explanation of the explanation, and probably to be found somewhere else.

Could be as simple as the difference between the open circuit voltage on one half with no train present, and the maximum volt drop on the other half with a train in it. I suppose there may also be regeneration to allow for, I can’t remember what the exact figure is, but I know the nominal 750 volts on our third rail can go well over 100 volts higher than that in extremis. I expect it’ll be a similar problem with overheads.

 

[eldomtom2]

Could be as simple as the difference between the open circuit voltage on one half with no train present, and the maximum volt drop on the other half with a train in it. I suppose there may also be regeneration to allow for, I can’t remember what the exact figure is, but I know the nominal 750 volts on our third rail can go well over 100 volts higher than that in extremis. I expect it’ll be a similar problem with overheads.

The standard voltage for Japan's overhead DC system is 1500V.

 

[swt_passenger]

The standard voltage for Japan's overhead DC system is 1500V.

Yes I understand that, but a pro-rata variation between no load and full load will still be possible, so you could even see a variation of around 200 volts.

 

[norbitonflyer]

In 3rd rail land, all the shoes in a unit are connected and it is quite possible for two sections to be bridged by a unit. (There has been at least one incident where an alert guard noticed that the compressors on the unit he was detraining had started up, indicating that the supposedly dead section his train was in was actually live. The cause was another train which had been allowed to overrun the end of the adjacent live section and was feeding current to the dead one.

 

[contrex]

A carbon arc lamp needs less than 100V between the ends of the carbon rods (usually a few mm apart) to maintain the arc

When I were a lad, arc welding kits were available for home use that required you to supply a car battery (12v of course). A quick Google shows that they are still a thing. I remember a bloke on telly years ago who reckoned he knew how ball lightning was generated, and he had rigged up an old submarine battery (A much higher voltage, I would imagine, typically 200-400 volts, and thousands of amps potentially) with a momentary shorting pendulum switch. It made exciting bangs and fizzes and shot out luminous ball-shaped blobs. I hoped he had adequate fire insurance.

 

[stuving]

I've found some French/Belgian documents about DC OLE practice, and had a look at Garry Keenor's book for comparison and English names. I suspect the way DC OLE is built is pretty standard, but how it's operated may vary a bit. Obviously this is mostly familiar from 25 kV, but not wholly.

French practice is for each feed station and mid-point (and any other sectioning pints) to have a break in the OLE lines, with switches to connect the two sides and paralleling switches between the tracks. Breaks are single: neutral sections are not usual. The breaks are of two kinds, a section insulator or an insulated overlap.

A section insulator (isolateur de section) usually has a short "make before break" overlap, though Arthur Flury make a few that are labelled as "with neutral section". I was expecting to find a lot in makers' technical data about management of arcs, or whether crossing the break under power has to be avoided - but it's not mentioned. An insulated overlap (sectionnement à lame d'air) is the same as the "air section" in the OP.

But how this installed kit is used is less clear. I can think of several possibilities for the switches between sections:

1. They are always closed, unless a section of OLE is to be taken out of use. That means no electric trains (or noen at all), and for work on the OLE some complex rules about grounding and leaving the next section out of use as well (to create a neutral section). I think that applies to France.
2. They are always open, unless a feed station is out of use and feed from the next section(s) is needed.
3. They are usually closed, but open often enough that driving procedures must be compatible with crossing a "live" section break.
4. etc.; there are other possibilities.

 

[edwin_m]

I've found some French/Belgian documents about DC OLE practice, and had a look at Garry Keenor's book for comparison and English names. I suspect the way DC OLE is built is pretty standard, but how it's operated may vary a bit. Obviously this is mostly familiar from 25 kV, but not wholly.

French practice is for each feed station and mid-point (and any other sectioning pints) to have a break in the OLE lines, with switches to connect the two sides and paralleling switches between the tracks. Breaks are single: neutral sections are not usual. The breaks are of two kinds, a section insulator or an insulated overlap.

A section insulator (isolateur de section) usually has a short "make before break" overlap, though Arthur Flury make a few that are labelled as "with neutral section". I was expecting to find a lot in makers' technical data about management of arcs, or whether crossing the break under power has to be avoided - but it's not mentioned. An insulated overlap (sectionnement à lame d'air) is the same as the "air section" in the OP.

But how this installed kit is used is less clear. I can think of several possibilities for the switches between sections:

1. They are always closed, unless a section of OLE is to be taken out of use. That means no electric trains (or noen at all), and for work on the OLE some complex rules about grounding and leaving the next section out of use as well (to create a neutral section). I think that applies to France.
2. They are always open, unless a feed station is out of use and feed from the next section(s) is needed.
3. They are usually closed, but open often enough that driving procedures must be compatible with crossing a "live" section break.
4. etc.; there are other possibilities.

Not quite the same, but I did some testing on the Croydon tram system before opening so got pretty familiar with the design. Every substation could feed one or other direction or both, but normally fed both so the two sections were electrically connected via a common busbar. In this way the whole network was normally in parallel. Breakers could be opened to isolate the section from one substation to the next one.

This allowed a fun experiment where we earthed the OLE at one end of the Birkbeck-Beckenham section and connected a big signal generator to the other end, so we were circulating quite a few amps at different frequencies out along the OLE and back along the rails, so as to measure the induced current in the parallel heavy rail section (under possession I hasten to add). The OLE made a nice singing noise as the frequencies were changed.

 

[ac6000cw]

In this way the whole network was normally in parallel.

Do you know if that is normal practice with tram/light rail systems?

It's interesting because in the RAIB report on the T&W Metro incident mentioned in post #4 it shows and talks about that system having 'section gaps' and the train involved coasting through a neutral section (see sections 17 & 30 of the report).

 

[507 001]

Do you know if that is normal practice with tram/light rail systems?

It's interesting because in the RAIB report on the T&W Metro incident mentioned in post #4 it shows and talks about that system having 'section gaps' and the train involved coasting through a neutral section (see sections 17 & 30 of the report).

Yes it is normal practise.

‘Neutral sections’ aren’t a thing on most of the DC OHLE systems I know of. On Metrolink sections are divided by Section Isolators or SIs, which are live throughout.

 

[swt_passenger]

Yes it is normal practise.

‘Neutral sections’ aren’t a thing on most of the DC OHLE systems I know of. On Metrolink sections are divided by Section Isolators or SIs, which are live throughout.

Perhaps it’s because of the junction, and the stabling sidings and the (not shown) depot avoiding line. If you got into a situation where a section of track was effectively fed from 3 locations, perhaps the fault monitoring and tripping system would get too complicated?

 

[507 001]

Perhaps it’s because of the junction, and the stabling sidings and the (not shown) depot avoiding line. If you got into a situation where a section of track was effectively fed from 3 locations, perhaps the fault monitoring and tripping system would get too complicated?

I’m no electrical engineer, but I’m fairly certain there are places on Metrolink where just that situation occurs (Irk Valley springs to mind, a potential for up to 4 feeds there). Still no neutral sections.