Thread by @25kV

THREAD: A few of you requested a #railwaysExplained thread Return Conductors (RCs) and Auto Transformer Feeders (ATFs). As part of that I'll be attempting to explain how immunisation works. HEALTH WARNING: this will involve electromagnetism. Apologies in advance!

1/

Lets start with the very basics: overhead line forms part of an electrical circuit. Just like all circuits, the electricity flows out from a supply (the feeder station) to the load (the train) and then flows back to the supply. The OLE forms the outward leg of the circuit

2/

The simplest way to form the return half of the circuit is simply to use the running rails. Connect the non-OLE side of the train motor to the wheels, current flows through the wheels and into the rail, then back to the feeder station.

3/

Why does this matter? Well if you want to generate an EM field, all you have to do is pass some electricity down a wire. Hey presto, an EM field is set up around the wire.

6/

If the generating circuit is AC, this field can then interact with other electrical circuits *even if they aren't physically connected*. They can be miles away, as your FM radio demonstrates every time you switch it on.

7/

The strength of an EM field generated by a circuit is proportional to the amount of current that flows. So when ~200A flows in OLE, it generates a BIG field. So what? Well, OLE exists in the real world, with other electrical circuits in close proximity. Like signalling cables

8/

Signalling cables carry small voltages & currents. When an EM field cuts cross this electrical conductor, it induces a voltage in the wire according to Lenz's law. This says that the induced voltage is proportional to the strength of the field AND the length of the conductor

9/

So we have a big EM field and a long signalling cable. We therefore get a large induced voltage; a 200A train load with 10km of parallel lineside cable will induce a longitudinal voltage of 154V in the cable.

Congratulations, your red signal just turned green. Not good.

10/

So when railways began to electrify with AC in the 1950s, these new circuits started interacting in unpleasant ways with not just signalling cables, but radio antenna (famously, Jodrell Bank), hospital equipment, etc etc. Something had to be done.

11/

The solution that was devised uses another useful feature of EM fields; they are vectors - in other words, they have direction. Things with direction can be added & substracted. If we have two EM fields of the same strength but opposing directions, they cancel each other out

12/

So if we placed another conductor carefully, and passed an equal and opposite current through it, at any point equal distance from the two conductors, there would be no EM field.

13/

The second conductor is called the Return Conductor (RC). It is positioned on the back of the OLE mast, at the same height as the OLE, and so roughly same distance away from lineside cables. Only problem is, how do we get an equal and opposite current to flow in it?

14/

The obvious place to get that current is from the running rails. After all, the current is flowing back to the feeder station (opposite direction to OLE) and should be roughly the same number of amps.

But we have to move all that current from the rail to the RC. How?

15/

Enter the Booster Transformer (BT). This is a current transformer - for every 1 amp that flows in one winding, 1 amp is forced to flow in the other winding. With a nice irony, this transformer uses, you guessed it, electromagnetism to produce this effect.

16/

All that remains is to place the BT into the traction circuit in such as a way as to force the outgoing current to flow through one winding of the BT, and connect the RC to the other winding. Then connect the running rails to the RC.

This is known as the BT/RC system.

17/

As the train demands current, it flows through the BT winding. An equal and opposite current begins to flow in the secondary winding and thus along the RC. There is only one place this can come from - from the train, through the rails.

18/

Now the (roughly) same number of amps are flowing in OLE & RC, but in opposite directions. The EM fields largely cancel at ground level, and so our signalling cables will see a much-reduced induced voltage. Our 200A train with 10km of cable will now induce only 12V

19/

The RC will have a voltage due to volt drop in the system - that is why RCs are always mounted on insulators.

That's it for BT/RC… next up: Auto Transformer Feeding

20/

THREAD EXTENSION: here we go with auto transformer feeding. Buckle up, this is a messy circuit diagram!
/21

The BT/RC system is part of a 25kV distribution system. There is a limit to how much power any electrical system can deliver, and supply voltage and circuit impedance (resistance) are the two key constraints. Booster Transformers increase system impedance, which isn't good
/22

So how can we deliver more power? It is hard to reduce the impedance of the system without making the OLE conductors large and heavy, and the trains must always see 25kV, so voltage increase is out...

Well, maybe not, if you use a very clever circuit design.
/23

What if we could distribute power at a higher voltage - say 50kV - but configure the system so that the train only sees one half of the circuit? Power is proportional to V^2, so theoretically you get 4x the power.

This gets a bit complicated but I'll do my best to explain
/24

I like to visualise a simplified circuit like this. If you can insert a potential divider (PD) in the circuit, and feed it at 50kV, and each half of the potential divider has the same impedance, then the load (train) will only see half the voltage.
/25

But this isn't an ordinary PD; each half of the PD is a transformer winding, so by setting the transformer up correctly we can force equal and opposite current to flow in each winding. We're effectively transferring power from one half of the winding to the other.
/26

This special PD is an Auto Transformer (ATx). This is a current transformer, like the BT. But unlike the BT it has centre-tapped windings - the two windings are connected at one end and there are three connection points to the outside world rather than the normal four.
/27

At around 5km intervals along the route, a lineside ATx is provided which acts as a mini feeder station. It receives power at 50kV - from OLE & ATF - but transfers power from ATF to OLE circuit via the windings, and feeds that power to the OLE at 25kV.
/30

This system is a mesh, with multiple circuits; which makes it hard to understand if you're not an electrical engineer. But as the diagram shows, each ATx provides part of the current that the train is demanding.
/31

The big advantage of AT feeding is that you can run more trains, or higher power trains, or space the feeder stations further apart and deliver the same power. Average AT feeder station spacing is 80km; whereas classic feeder station spacing is 40 to 60km.
/32

The downside of AT feeding is that additional live conductor. Getting it through stations and bridges often requires cable terminations, insulated cables and trough route. Which isn't cheap.
/33

We've seen up thread how the Return Conductor (RC) needs a Booster Transformer (BT) to provide full immunisation, but that BTs require maintenance and limit the power output of the system. So recent years has seen Network Rail make a concerted effort to replace BTs.
/33

This system is known as Boosterless Classic Feeding. There are no BTs or RCs in this system - we're back to a simple rail return.

But you still need to suppress electromagnetic interference, so how else can you do that? Well, you exploit Lenz's Law once again.
/34

MSCs and RSCs are insulated aluminium cables, laid at ground level, usually inside the concrete trough route carrying the signalling cables that you want to protect.

If it is an MSC that you want, you connect this cable to earth every 1km with earth rod(s).
/35

How does it work? Well the EM field generated by the OLE acts on the MSC, and sets up an opposing magnetic field as per our friend Lenz. Current then flows in the MSC, in the opposite direction to that in the OLE.

With me so far?
/36

The same effect also happens in the signalling cables, but the cross sectional area of the MSC is much larger than that of the signalling cables, so the current is correspondingly higher too.
/37

The MSC current creates its own EM field, which induces a voltage in the signalling cable - which drives current in the opposite direction to that directly induced by the OLE. This works as long as the MSC has low impedance to earth and is very close to the signalling cable.
/38

A typical 19/3.25 Aluminium MSC provides around 31% reduction in induced current compared with no measures. Depending on the specifics of the system, this can be enough.
/39

BUT there is a better option - one which needs no expensive and vulnerable earth rods AND provides more immunisation - the RSC. An RSC is the same cable, but is not connected directly to earth. Instead it is connected to the traction bonding system every 400m.
/40

By doing this, the reduction in induced current is increased to around 66%.

RSCs are now the standard immunisation mechanism for new electrification in the UK; this will be the case until copper signalling cables are replaced with fibre optics at some point in the future.
/41

That's it for this epic thread - sorry it took so long to finish! You can read all my other #railwaysExplained threads here:
/42

https://twitter.com/25kV/status/1187822175010590721

And I should add - you can read more about all of this in my free e-book at section 9.4ocs4rail.com

THREAD CORRECTION: when I wrote this about the Auto Transformer system, I was using a circuit diagram that I now know to be incorrect. (In my defence, so were SNCF).

These are the correct diagrams

Also - auto transformers are NOT current transformers as I previously stated - they are auto transformers, and Very Weird Indeed, with a portion of the primary winding also forming the secondary. WTF?

If you want to try to understand them fully I suggest you have a read of en.m.wikipedia.org/wiki/Autotrans…