It was never a good time for a train delay. After a busy day in the city, I didn’t have anywhere to be except home for a late dinner and an unmoving night on the sofa. But I was too exhausted to cope with any deviations from that plan, so I groaned when the train stopped just before one of the stations on the way. A red light, I assumed, until I looked out the window and froze in my seat. The wires of the overhead line swung dangerously back and forth, still intact, yet caught in such a terrible gust of wind that the question didn’t seem to be if they’d break, but when.
A shudder went through the train. In spite of my fear, I noticed the sky: it had turned deep purple, orange, and yellow, with only a single row of low clouds, and even those seemed to be rushing away from this place. Another shudder, and a shadow fell. The shadow of a mighty, pointed wing—and as a giant dragon flew just above the treetops not so far from the train, its gold-and-purple scales continued to reflect the light into the most beautiful sunset I had ever seen.
Perhaps this view was worth the delay.
—
[Image description: Photo of a stunning sunset above the black outline of some houses, trees, and street lanterns. Orange sunrays burst through a low line of clouds into the deep blue, almost purple, of the sky above. The black outline of the railway’s overhead lines is at the front of the picture.]
—
Adding a little announcement: the Kickstarter for The Neurodiversiverse: Alien Encounters will launch soon—this is an anthology of sci-fi short stories, poetry, and art, in which a short story by Paranatellonta writer Minerva Cerridwen will be published! Click “notify” on this page to learn more. Thank you!
Let’s talk about railway electrification systems; not so much because any of you care but because I want to talk about it (and because I talked with some people about it on Twitter recently and had a lot of fun with that).
There are four main voltages that overhead lines for european railroads can have (and most of this translates to other countries as well):
1500 V DC
3000 V DC
15000 V AC at 16.7 Hz frequency
25000 V AC at 50 Hz (aka normal) frequency
These are generally divided along country lines; every country has one (or somtimes two) of these that it uses, usually chosen without any regard for what their neighbours used. Why?
Well, because back at the start of the 20th century, everyone had different ideas and different trade-offs and different decisions, and those choices made sense to them at the time. Many of these factors have changed since; for example, after World War 1, France explicitly decided on 1500 V DC so that german locomotives (running on 15 kV AC) could not run into the country. It made sense then, it’s a pain in the butt now.
But why those values in particular, and what were the tradeoffs?
The Motor
The key for that is the “Universal Motor” (for my german-speaking followers: Einphasen-Reihenschlussmotor), a type of electric motor that was the standard for electric railways and many other applications from the first electric locomotives until 1980.
I’m not going to go into details; the important part is that this motor is essentially a DC motor, but due to its wiring, it can also run on AC. There’s a bit of an issue, though: Powerful universal motors don’t really run well at industrial frequencies (50-60 Hz); they prefer lower ones. Other than that, the motor is bulletproof and powerful and easily the best thing that doesn’t require computer control.
Your standard railway spec motor, small enough to fit comfortably into the running gears of a train, will take somewhere between 600 V and 750 V (with quite a wide margin at either end). To this day, most streetcar systems, subways, and some odd commuter rail lines (in particular most of the south of England) use just those 600-750 V directly, as DC current, because the motor runs better that way.
More Power
But low voltage means low power. Power is voltage times current, and more current means generating more heat in the overhead power lines. For a given level of power, having more volts means needing less current, which is better for the power lines. And any design of power line will have a hard limit of how many amps of current you can send through it before it’ll melt. You can build lines for more current, but that’s expensive.
The low power does not only mean that each individual train gets a low amount of power, but also that the number of trains in a given section has to be low. So you need a lot of power supply stations (substations).
So the first idea was to connect two universal motors in series. That takes 1500 V, and you get either twice the power at the same current, or need half the current for the same power. That is what is used for example in the Netherlands, France and Japan.
1500 V is still fairly low, though, so why not double that? For 3000 V DC, you connect all four motors in your typical four-axle train car in series. Again, more power, less current. That system is used in Belgium, Spain, Italy, Poland, Czech Republic, Slovakia, former Soviet Union and a couple of others.
You can’t really go beyond that, though, at least with 1920s tech. Your modern USB charger is actually an incredible feat of engineering; it took decades to reach that level. At the time, if the DC voltage you got out of the power station didn’t match your needs, then there was no easy solution.
Transformation Sequence
This was not true for AC power, where you can use a transformer, an incredibly simple piece of technology. With AC, you can essentially use as high a voltage as you want. The limit here is insulation: The higher the voltage, the more space you need to have between the roof of the locomotive and the wires, and between wires and bridges and so on. The European countries that went that route settled on 15000 Volts as a good compromise.
The problem with that is that the universal motor doesn’t like 50 Hz (or 60 Hz) frequency that you get from the normal grid. The solution is to run the whole thing with less frequency. That’s why the frequency in the line is 16.7 Hz (originally 16 2/3rds Hz). This system, 15000 V AC at 16.7 Hz, is used in Germany, Switzerland, Austria, Norway and Sweden (but notably not Denmark), and it has stood the test of time well. For the Americans reading this, the 12 kV at 25 Hz used in the north-east by the Pennsylvania and Reading railroads is essentially the same thing, just slightly different values.
The problem with this thing is that you absolutely need that transformer. And, for reasons that I don’t quite understand, the lower your frequency, the larger your transformer has to be. 16.7 Hz is fairly low, so you need a very heavy transformer. Compared to a low-voltage DC system, you need fewer substations and a less expensive overhead line, but you need more expensive and heavier locomotives. That is a very real trade-off: Many of the DC countries have a long history of small, quick EMUs, while it took much longer for AC systems to develop those; they required heavy locomotives or much heavier EMUs.
Trade-offs
In the 1920s and 1930s, when the foundations of modern electric networks were laid, these were the systems and considerations available, and given the train performances at the time, it’s hard to argue that anyone really choose wrongly. I keep saying that 1500 V means low power, but the french reached a world record for high speed trains at 331 km/h (a bit over 200 mph) in 1955 with that system.
High Frequency
In the 1930s, hungarian engineer Kálmán Kandó, already an incredibly important figure in the development of modern electric trains, developed phase converters for railway use, which changed the game completely. These things were at the time heavy mechanical devices that combined a motor and a generator, and could transform any sort of electricity into any other. This means you can use the normal 50 or 60 Hz power that comes from the national grid, and then transform them into something else in the locomotive, instead of using some weird 16.7 Hz thing. This means cheaper lineside equipment and smaller transformers. You need the converter, but that pays off almost immediately.
(I'm over-simplifying here; there is a number of technologies and types of motors that allow using 50 Hz; the phase converter was the first, but is far from the only.)
Railways at the time were very interested in that, but then there was a whole second world war, which put everything on hold for a while. After the war, engineers in various countries perfected it, and along bumped up the voltage to 25000 V for more power with not that much more insulation required (the one exception to this is Japan, which went for 20000 V instead). This 25 kV at 50 Hz (or 60 Hz where applicable) is nowadays generally considered the best system if you can choose freely.
All countries that have 1500 V or 3000 V also have more or less extensive networks with 25 kV 50 Hz; sometimes just for high-speed lines, often for about half the country. A number of countries that started electrification comparatively late only have 25 kV 50 Hz. The countries that already have 15 kV 16.7 Hz have stuck with it, though; 25 kV 50 Hz is better, but not so much better that it justified all the expense of adding a new system.
(Exceptions exist but are very rare; feel free to ask me about the Rübelandbahn.)
Modern Locomotives
On the locomotive side, things have drastically changed starting in the late 1970s. Thanks to modern microelectronics and modern power electronics (sadly nobody calls them macroelectronics), phase and voltage converters have become small, lightweight and incredibly versatile; they’ll convert anything to anything else and back if you design them properly. That makes it relatively simple to build a locomotive or EMU that can use all of these different voltages, something that used to be quite a major engineering feat.
The default approach as of right now is that you have a big transformer (no way around that, for now) for AC voltages, with different output settings for 15 and 25 kV (this was always easy). It has to be the 16.7 Hz size, sadly. The output voltage in the 1500-3000 V range gets turned to DC. If you’re running under DC, you just use that DC directly. No matter how you got that DC, you’re now putting it into another converter (typically one per wheel set or one per bogie), which turns it into three-phase AC, at a frequency that corresponds to the speed you want to go. This sounds complicated, but works well in practice, to the point that all new locomotives nowadays support either both 1500 V DC and 3000 V DC, or both 15 kV AC and 25 kV AC, or all four. You can’t get a new e.g. 15 kV AC locomotive anymore. Even ones that are designed just for one country and advertised as doing only 15 kV will actually be able to run with 25 kV, just because nobody thought it worth the effort to design a 15 kV only transformer.
(This is not true for EMUs, since those are designed to run mostly locally instead of through the entire continent. You can get them in multiple voltage, but most are designed for just one.)
Where do we go from here?
The current system is a mess that is interesting to me, but a bit of a problem for railroads. In the olden days, you’d just change the locomotive at the border anyway, so it didn’t matter much; but nowadays you want to run your freight train from Rotterdam to Genua non-stop if you can. While multi-system locomotives have become a lot cheaper than they used to be, the whole thing is still very annoying for cross-border traffic. It's not the only annoying thing about european cross-border rail traffic, but it's a factor.
Also, train sizes, speeds and power requirements have increased drastically (air conditioning in passenger trains is actually a big deal in terms of power use). AC systems have been able to cope; DC systems less so. Both Belgium and the Netherlands have quite a lot of diesel locomotives pulling freight trains on electric lines, because they can simply produce more power. One manufacturer offers electric locomotives (the Stadler Euro9000) that have a diesel motor to boost the power under DC (and for shunting on tracks that have no overhead lines). Clearly, the old DC system needs to go.
But that’s easier said than done. Converting a line to 25 kV is quite expensive. You need new insulators at every single overhead line mast, but even more importantly, you need to check for safety clearance at every bridge over the railway, and, if necessary, raise bridges or lower tracks accordingly. Tunnels get even more fun. And, of course, all line side power equipment needs to be swapped out. This isn't impossible, but it is very expensive, and while it has happened in some places, it hasn’t happened a lot.
The Netherlands, for example, are currently running on 1.5 kV DC and did consider switching to 25 kV AC. All modern equipment there actually has a bit of empty space where you can fit a transformer. But the cost of upgrading the lines was judged too high. Right now ProRail, the company in charge of the network, is proposing upgrading from 1.5 kV to… 3 kV DC, of all things. It seems like a minor deal, but it still allows doubling the power output, for much less cost. Will it happen? No clue.
Meanwhile, in France, there is research going to make use of the new electronics. Researchers there figure that 9 kV DC is something you can do very well; modern electronics should allow stepping that down even more efficiently than a big 50 Hz transformer could. This seems to be mainly because the government does not want to pay to change the 1500 Volt lines there to 25 kV.
On the other end of the spectrum, the topic of 50 kV AC keeps popping up every now and then. A few lines like that already exist, most notably a large one in South Africa. In the US, the Black Mesa and Lake Powell railway used this system and ran coal trains from a mine to a power station; it's closed since 2019, because the power plant closed, because shale gas and renewables are just plain cheaper than coal. It's an interesting bit of railway history lost, but definitely a net win for the planet.
From a technical point of view, there's nothing particularly difficult about this. The Black Mesa and Lake Powell actually used a number of locomotives built for 25 kV and just changed the windings in the main transformer. The big problem is safety distances, which are much bigger than for 25 kV. There is no mainstream push for 50 kV at the moment, but it keeps coming up in discussions about "should US railroads electrify their lines" as a possibility.
Finally, in the 15 kV countries, there is some thought about 15 kV at 50 Hz. In Germany, the idea is to use this for short recharging sections for electric trains with batteries. Using 50 Hz saves the expense of a frequency converter. Personally, I don't see why those sections couldn't be 25 kV, but I guess it makes things a bit cheaper (EMUs, unlike locomotives, still come in 15 kV only versions).
In the US, this is actually already a thing; some lines in the North-East (I think primarily for New Jersey Transit) were changed from 12 kV 25 Hz to 12.5 kV 60 Hz when the 1920s era line-side equipment needed replacing. For the newer trains there, this requires at most a software update.
So… that’s the current situation. It’s not likely to get better any time soon, and if the french 9kV DC plans go through, it may actually get a bit worse, but modern locomotive technology has evolved to cope. There’s no point to this post, I just think it’s fun.
