Power systems engineering must be fun

[IanB]

Just looking at the new Eurostar e320 train sets: https://en.wikipedia.org/wiki/British_Rail_Class_374

They can develop 16 MW (21,000 hp) and run at 320 km/h (200 mph).

Their power supply can be any one of the following:
- 25 kV AC 50 Hz
- 15 kV AC 16.67 Hz
- 3000 V DC
- 1500 V DC

Though presumably they can only develop the maximum 16 MW on the high voltage AC supplies.

How would this be designed? Presumably the incoming supplies are all converted first to a common DC power rail, from which the variable frequency drive for the motors is operated. But with two completely different AC frequencies, transformers to reduce the incoming voltage are going to be interesting. And does the 1500 V DC get stepped up, or does everything work off a 1500 V power rail? Because that would be in the order of 10,000 A for the required power.

The whole thing seems like a fascinating piece of electrical engineering.

If I were going to study EE, this is the kind of project I would like to work on.

[Zero999]

I wouldn't have thought it would use a transformer to convert the AC voltages to a lower DC voltage. It's far more likely the AC is rectified and converted to a steady DC voltage with a switched mode converter.

[CM800]

Just looking at the new Eurostar e320 train sets: https://en.wikipedia.org/wiki/British_Rail_Class_374

They can develop 16 MW (21,000 hp) and run at 320 km/h (200 mph).

Their power supply can be any one of the following:
- 25 kV AC 50 Hz
- 15 kV AC 16.67 Hz
- 3000 V DC
- 1500 V DC

Though presumably they can only develop the maximum 16 MW on the high voltage AC supplies.

How would this be designed? Presumably the incoming supplies are all converted first to a common DC power rail, from which the variable frequency drive for the motors is operated. But with two completely different AC frequencies, transformers to reduce the incoming voltage are going to be interesting. And does the 1500 V DC get stepped up, or does everything work off a 1500 V power rail? Because that would be in the order of 10,000 A for the required power.

The whole thing seems like a fascinating piece of electrical engineering.

If I were going to study EE, this is the kind of project I would like to work on.

It is an interesting area to work / experiment in. It's the kind of project that originally got me into power electronics and to the job working with servo drives that I do now. I mostly have interest in high voltage power supplies etc. the theory is very simple, it's just scaling up the siz and adding protecgtions.


Reading Wiki, the 1500V is overhead line. I'd imagine the lower voltage (3000V and 1500V) are for inner city, it won't be drawing anywhere near as much as the full spec of the train as the trains can't got fast in crowded towns / cities.

I would indeed imagine the power train to run around 1500V using high voltage IGBTs, at most 3000V or so. Just check out the infineon IGBTs and some of the other power semiconductor manufacturers.

The 25kV and 15kV wouldn't be stepped down by a traditional transformer I bet, likely more solid state like they use for HVDC power lines. a 16MW iron core transformer for 50Hz would be big as it is, let alone 16.67Hz.

16.67Hz is a strange frequency for power... You'd need a dedicated powerstation / genset. No clue where they'd get it from otherwise.


Here we go:

The N700 Shinkansen uses a three-level converter to convert 25 kV single-phase AC to 1,520 V AC (via transformer) to 3,000 V DC (via phase-controlled rectifier with thyristor) to a maximum 2,300 V three-phase AC (via a variable voltage, variable frequency inverter using IGBTs with pulse-width modulation) to run the motors. The system works in reverse for regenerative braking.

[IanB]

I can understand the problem with the size and weight of iron core transformers, which means that rectifying and down-converting 25 kV AC must in itself be an interesting design exercise.

The 16.67 Hz apparently is in Germany. I guess they must have (had) dedicated power generation for the railways there, or maybe in modern times they derive it from the standard power grid for backwards compatibility.

[CM800]

Here we go, have a datasheet 

https://library.e.abb.com/public/a2113a079c793d71c1257a040031f3d1/PDS%20BORDLINE%20CC1500_AC_15-25kV_M_1500_3BHS296808%20ZAB%20E01%20Rev%20C.pdf

[IanB]

Interesting. So it does look like step down transformers are used for the incoming AC. I'm pretty sure transformers are used for regular trains on British railways because you can hear them humming when the train is in the station.

That datasheet shows a cabinet mounting arrangement, however, the Siemens Eurostar trains do not have dedicated drive cars, all the equipment is distributed throughout the train (presumably much of it underneath). That must also make for an interesting packaging exercise.

[Tomorokoshi]

Here we go:

The N700 Shinkansen uses a three-level converter to convert 25 kV single-phase AC to 1,520 V AC (via transformer) to 3,000 V DC (via phase-controlled rectifier with thyristor) to a maximum 2,300 V three-phase AC (via a variable voltage, variable frequency inverter using IGBTs with pulse-width modulation) to run the motors. The system works in reverse for regenerative braking.

I rode on an N700 once.

The experience is something like, "hmm, this isn't going very fast..." until one looks out the window, and realizes that the blur distance is about 3x out from where it would be with a car on the highway. They are very smooth all around.

I don't recall the sound of the motors on the N700. Probably because they were so quiet.

Now, for the city trains, one can definitely hear distinct frequency ramp ups with what almost sounds like a 3-speed transmission, where perhaps additional coils on the motor are getting switched in when getting up to speed.

Typical trains seem to be 16 cars long with a pantograph about every 4th car.

[CM800]

Interesting. So it does look like step down transformers are used for the incoming AC. I'm pretty sure transformers are used for regular trains on British railways because you can hear them humming when the train is in the station.

That datasheet shows a cabinet mounting arrangement, however, the Siemens Eurostar trains do not have dedicated drive cars, all the equipment is distributed throughout the train (presumably much of it underneath). That must also make for an interesting packaging exercise.

I doubt the new train uses transformers. It would have to be a multiple stage converter like the N700 Shinkansen

[IanB]

I doubt the new train uses transformers. It would have to be a multiple stage converter like the N700 Shinkansen

But note what you quoted about the N700:

The N700 Shinkansen uses a three-level converter to convert 25 kV single-phase AC to 1,520 V AC (via transformer) (stage 1) to 3,000 V DC (via phase-controlled rectifier with thyristor) (stage 2) to a maximum 2,300 V three-phase AC (via a variable voltage, variable frequency inverter using IGBTs with pulse-width modulation) (stage 3) to run the motors.

The ABB datasheet you linked to also mentions tranformers:

Incoming power from the catenary (15/25 kV) is stepped down by the transformer to feed the Compact Converter BORDLINE® CC1500 AC...

So all examples of technology we have seen in this thread use a step down transformer for the incoming AC. While it's possible the Eurostar train does not, we do not have information about this.

[radar_macgyver]

They all use transformers for the first stage (even N700). Even the ones that run on 16.67 Hz. Apparently the reason for that frequency is for compatibility with trains that ran on DC with series-wound motors (aka universal motors). At 50 Hz, the core losses were too high, so they chose to use 16.67 (which is 50/3 - generated using a motor-generator set, with the generator having only two poles while the motor had six). Obviously this means a HUGE transformer that the trains all lug around - which they'd rather not change because of all the existing infrastructure.

The really amazing ones are those that run direct off a 50kV line, like Sishen–Saldanha. Most "25 kV' systems are actually 50 kV balanced, with the track serving as the neutral and autotransformers along the line to maintain balance.

[NiHaoMike]

Don't train engines need to weigh a lot in order to get the required traction? The bulk of the 16Hz transformer might well be a feature.

[T3sl4co1l]

Not really.  Power systems is largely boring technology:

Transformers. Tons of big, dumb iron.

Slow IGBTs.  They take microseconds to turn on and off.  You don't even have to worry about stray inductances, just slap some copper bars on the things and you're good to go.

Very slow development cycles.  It takes years, decades even, to first get things right, and test it in the lab, and decades more to finally deploy.  All the while, the large sums of money involved are distributed over the same time scales, to lessen the blow.

Very long product lifetimes.  A new engineer might be responsible for over a half century of aging facilities, plus the few new and interesting things.  Once you've done all the cool work, you still have to maintain the system, maybe for the rest of your career!

Controls are interesting, but the large part is more about logistics and coordination: what safety features are necessary, how can the system be managed by the central database, etc.

If you wanted some alternative perspective, that is.

Tim

[Gromitt]

Just for information, here is a map of the different electrical systems for railways in Europe



All is overhead except for 750V DC in England.

/stefan

[Nauris]

Don't train engines need to weigh a lot in order to get the required traction? The bulk of the 16Hz transformer might well be a feature.

Sure, if you have traditional engine and separate cars unlike these high speed passenger trains.

[Galenbo]

Just for information, here is a map of the different electrical systems for railways in Europe

That's not a map of Europe, looks more like something in the direction of eurasia.

[Gromitt]

Just for information, here is a map of the different electrical systems for railways in Europe

That's not a map of Europe, looks more like something in the direction of eurasia.

There is not a lot of Asia in that map.

/stefan

[rstofer]

A brief discussion of the differences between AC and DC traction systems:
http://www.republiclocomotive.com/ac-traction-vs-dc-traction.html

[dr.diesel]

Individual motors can get quite big.  This one is used to lift water into cooling towers at a power plant (there is an identical one just to the left of it), 4000HP, but notice it makes that at 352RPM!

[CM800]

Individual motors can get quite big.  This one is used to lift water into cooling towers at a power plant (there is an identical one just to the left of it), 4000HP, but notice it makes that at 352RPM!

I bet that sounds lovely..... but honestly I don't think I'll find another motor that sounds as good as THIS:



Possibly:

[Red Squirrel]

It would be interesting to see how they switch large amounts of power within the power converters, just bigger mosfets and more of them?  Or  something more involved? 

[dr.diesel]

It would be interesting to see how they switch large amounts of power within the power converters, just bigger mosfets and more of them?  Or  something more involved?

No converter on the one I posted, you simply drop it on the 4kv line and let it eat.  An operator is staged at the motor, makes sure the area is clear, hides behind something and the control room closes the breaker!

Faults are not that uncommon, I've personally seen several blow up while starting.

[T3sl4co1l]

On the ones that do use converters: IGBTs are available up to 4600V and 3000A or thereabouts.  No problem using a sextet to run a motor (in PWM, or twelve for a tri-level converter with more efficient PWM or modified sine wave), or stacking a few to convert high voltage down to a useful DC link voltage that such a VFD can run on.

Tim

[T3sl4co1l]

http://www.infineon.com/dgdl/Infineon-FZ750R65KE3-DS-v03_01-EN.pdf?fileId=db3a304325afd6e00126461fd3936974

Correction, up to 6.5kV. This is either a smaller one (at only 750A), or straddles the curve of maximum capacity, where volts and amps trade off (which is logical enough, as die volume roughly corresponds to switching V*A capacity: thicker dies carry more voltage, wider dies carry more current).

The voltage drop is quite low, analogous to a 650V MOSFET that drops 0.45V at 7.5kA.

Tim

[tautech]

It would be interesting to see how they switch large amounts of power within the power converters, just bigger mosfets and more of them?  Or  something more involved?

No converter on the one I posted, you simply drop it on the 4kv line and let it eat.  An operator is staged at the motor, makes sure the area is clear, hides behind something and the control room closes the breaker!

Faults are not that uncommon, I've personally seen several blow up while starting.

When I started work as a teenager there was a 250 HP driving a huge chipper, the faceplate of which must have weighed more that a tonne and the startup was controlled by a worm driven staged knife switch, all immersed in an oil bath about half the size of a 44 gal drum. When the last phase kicked in all the lights in the mill dimmed and the surrounding area shook as see got up to speed. IIRC the start process was ~20 seconds.
Only worked there for a few months but that memory lingers. 

[rrinker]

 The oddball frequency is a legacy thing, also employed in the US in the Northeast Corridor (former Pennsylvania Railroad). The low frequcny AC makes it possible to use universal motors as traction motors with simple transformer tap speed controls - consider this was done originally in the 1920's and 1930's. There's some good information in this Wiki article: https://en.wikipedia.org/wiki/Amtrak%27s_25_Hz_traction_power_system

 In the 60's they built some new electric locomotives for these lines and used DC traction motors, the first ones use mercury arc rectifiers in them - but they often caught fire. GE built the locos and their name for the mercury arc rectifiers was "Ignitron" - whole locomotive catching fire was a foregone conclusion. Later advances in solid state electronics allowed a more successful locomotive that used silicon rectifiers.