Long-distance transmission of electricity

Factual Questions
41

Maybe the distribution system in your neighborhood was changed at some point.

As someone else has stated, not much has changed in transformer technology. A transformer’s physical dimensions are generally proportional to the KVA rating of the transformer although as we start dealing with higher voltages, we need to account for increased insulation requirements (which in turn, makes the transformer bigger). Let’s ignore that for now though and deal just with KVA rating.

I have a small green transformer on my front lawn that feeds my house and three or four neighbors. It converts 4,160V to 120/240 which is what we typically use in US residences. I’m guessing it’s maybe a 50KVA - 75KVA (50,000 - 75,000 watts) transformer.

Up around the corner is a fenced off substation (and an associated building) where a much larger transformer steps 13,200V down to 4,160V and feeds my entire neighborhood (i.e. all of the little green boxes including the one on my lawn) as well as other developments and local shops etc. That’s at least a 2MVA unit (2,000,000 watts) but I can’t get close enough to confirm. It’s maybe 8 feet tall, wide and long

That 13200V transformer is one of many that are sitting on a 35000V loop or maybe a 69000V or 138000V loop (depending on various factors). This loop would be at a county-level scale. The transformers used at this level are very large since they have the capacity to power a small town. 100,000 KVA (that’s one hundred million watts) is not unheard of. Something that size is approx 18 feet tall, wide and long and weighs about 75 tons.

So anyway, maybe at some point, the distribution system in your hood was changed. They could have just moved the big transformer to another location.

(To those reading who understand the difference between watts and volt-amps, I do too. It would just complicate matters though).

42

(emphasis mine)

I’m guessing this relates to the fact that the type of clocks that rely on the local AC frequency for timekeeping were once very common but are now fairly uncommon?

43

I’ve not seen one in years. I think the old style institutional clocks that would be up on the wall in older libraries and gyms worked this way.

The first obvious problem is that every time there’s a power outage, there went your time accuracy. The second obvious issue is that these days, the electronics for a clock based on an oscillator are going to be a buck or less, and the resulting device is going to be much more accurate.

The downside is that the electronics need batteries, running off 120 volts requires a power supply and ultra-capacitors for backup.

44

That did factor into the decision-making process.

When you connect a bunch of generators together, controlling their frequency gets to be a bit difficult. If you try to make one generator go faster, instead of it actually spinning faster, all it does is tries to supply more power to your “grid” of generators. Similarly, if you try to slow a generator down, it doesn’t actually slow down. All it does is supplies less power to the grid. Try to slow it down too much and it will actually become a motor, being powered by the other generators on the grid. But it won’t actually slow down (well, maybe a little, but not much).

If you want to speed up or slow down the overall grid frequency, it has to be a coordinated effort between all of your generators simultaneously. It takes a lot of effort and from a technical point of view, it’s a big pain in the backside.

It used to be that the power frequency in the U.S. was regulated fairly tightly. The frequency would sag a bit during the day as that is when the heaviest electrical loads are, due to things like air conditioning running at its peak since the days are hotter than nights, buildings having their lighting on during the day, etc. The power companies would then speed the grid up at night to catch up, so that over a 24 hour period, the frequency would average out to exactly 60 Hz. So all of those AC synchronous clocks would lose a bit of time during the day, but they would make it up at night, and from day to day they would not lose any time.

AC synchronous clocks have mostly gone the way of the dinosaur. And to be fair, if you do still have an AC synchronous clock and it gets off by a few minutes every month, it’s easy enough to set it to the proper time and fix it. So a lot of people thought that all of this huge effort and expense that was being used to keep the AC frequency accurate in the long term was a huge waste of effort, and a waste of money. But no one was really sure how many people really depended on the accuracy of the line frequency.

So, in 2011, they decided to do an experiment. Rather than keep the line frequency stable by speeding up the generators at night, they would let the line frequency drift a bit. They planned on doing this experiment for a year, just to see who complained about it.

I read news stories about the experiment before it happened, and then I haven’t heard anything since. I guess very few people complained, as nothing made the news. As far as I am aware, they never really “ended” the experiment and now the overall line frequency is now allowed to drift a bit, and nobody really cares. It sags a bit during the day when the load on the generators is the heaviest, and speeds up a bit at night just due to the load easing up a bit, but that’s it. It’s still roughly 60 Hz on average, but there is no coordinated effort to get it anywhere near as accurate in the long term like they used to.

45

I’m pretty sure I still have one somewhere - it used to be my kitchen clock. I didn’t like the constant hum of the motor, so I stopped using it.

46

I think the other thing is that electronic devices which care about the mains voltage and frequency (e.g. using linear power supplies) have pretty much gone the way of the dodo, along with the clocks. Practically everything I’ve encountered in recent years uses a switch-mode power supply which doesn’t give a rat’s round arse what you feed it. they’ll take 90-250 VAC at 50 or 60 Hz which covers pretty much every region in the world.

the building I work in still has a number of these throughout.

They haven’t worked in years.

47

This was true for AC clocks in general for a long time. :slight_smile:

The big dial clocks in my grade school weren’t very accurate. They were frequently off. Occasionally you’d see the clock hand advance quickly to catch up to the right time. This means there was a whole 'nother system going on.

Since the adjustment was only a catch-up method, no way of going back, to set the time back the clock was semi-frozen for a bit. You’d see the second hand quiver for a while. The old time-seems-to-standing-still-during-class thing came true.

For time changes, which happened during the Monday morning after the official change, the clocks would zip forward quite fast to catch up. Going the other way, it was just frozen for an hour.

48

I believe increasing the voltage on an induction motor does decrease the amount of slip required to produce the same torque, thus allowing the 1725 rpm motor (with 75 rpm of slip at full load and rated voltage) to run a little bit closer to 1800 rpm with the same load. However, it drives the motor further past the magnetic saturation region, which rapidly drives up current.

49

That was a master-clock system. Since accurate clocks used to be expensive and difficult, you’d have just one clock, with lots of clock faces synchronized to it. That’s particularly desirable for a school, where you want everyone to start, stop, come, go, and change classes at the same time. It’s supposed to stay in sync all the time, so you never see it catching up.

But schools always put in the cheapest system they could buy, and those systems /never/ worked. Even when new. And after a while, they worked even less.

50

The numbers have changed since 1980, practical distances are longer, and continue to get longer. The Chinese are putting in some very large grids at present. Also, the Sahara, although still a long way from Japan, isn’t that far from Italy. The problems with the Sahara are political, not electrical. That I know about there was an attempt to get a large solar farm built in North Africa years ago. I’ve heard about more recent attempts but nothing detailed: is there any part of North Africa stable enough to support electricity exports to Europe?

51

Funny you should mention clocks…

52

Morocco is ideally situated, and Tunisia a close second. Morocco is working towards being a serious power exporter to Europe, and they do have long term stability. Tunisia seems happy for the moment, but things were a bit bumpy for a while there. Algeria sits in the middle of the two, and has the most land, but not the close by sea links to Europe. Politics and stability is a long term problem across the region. Europe probably doesn’t wish to become beholden.

Desertec seems to be one umbrella initiative, but hasn’t got very far.

53

I was wrong, they still work. just that no one has bothered to set them to the correct time for years.

54

Ignoring the recent experiments in decoupled grid frequency, synchronized clocks are far more accurate than quartz, because the long-term average is from an atomic clock.

Until recently, I used an electric clock radio that was synchronized to the grid, probably using a simple zero-crossing counter (no motor, obviously). This was the most accurate clock I’ve ever had (aside from internet connected ones); it would be accurate for years at a time–going through several DST changes, I’d change only the hour, not the minute. Quartz clocks are usually only good to ~15 seconds/month.

I know I didn’t just get lucky with a really good oscillator, because it had a battery backup mode, and it would drift >1% with the power out. So it just had some crappy RC circuit instead of a proper oscillator (just good enough to make it through a few minutes of power outage).

The clock maintained its accuracy despite just about everything else on the clock failing (I’m sure due to decaying electrolytic capacitors and other components).

55

Technically you’re right. So you’re saying the power utilities were “counting ticks” and making sure the exact number of 1/60th of a second ticks per year add up?

Because that would be pretty damn accurate. You’d always be within 1/60th of a second if there weren’t power failures.

Now I’m wondering how on earth they synchronized this. If it was as simple as 1 big generator, you could just increase the fuel/steam feed to that generator when you’re “behind” in ticks and decrease when you’re ahead. A digital counter would be tied to the electric grid at a major point and give you this “ahead” or “behind” signal.

Obviously it would be smoother if it were proportional control, such that if you’re a little ahead the signal quantity is small and so on. You could build the electronics to do this using 1950s technology.

The trouble is that the electric grid isn’t one generator. It’s hundreds and hundreds. How did they coordinate them?

56

Previous thread.

57

Yep. Although the tracking is typically at a daily level, not annual. And it can get more than 1/60 of a second off; possibly even many whole seconds off.

The grid frequency tends to be tied to load–high load slows it down slightly. And the load is highest in daytime hours, so that’s when the clock is likely to be furthest behind. At night, when the load is lower, the utilities will overspeed the generators slightly to make up for missed cycles.

Of course this isn’t going to be perfect, so there’s a long-term average to keep track of. I don’t know precisely how it works, but you can imagine that if you are 10 cycles behind on day 0 due to imperfect daily correction, then on day 1 you target (86400 * 60) + 10 cycles instead of the normal count. So at any given hour or any given day you might be off, but there is always a tendency to get back to the atomically accurate number. I don’t care if my clock is off by 10 seconds, but I do care if it drifts by a second a day, because then I have to set it all the time.

I don’t know exactly how the utilities coordinate efforts, other than that it’s complicated, and that when it goes wrong it can go very wrong. One thing to note is that every generator is exactly synchronized to the grid, but they know how hard they’re “pushing”. Think of it like a many-rider bicycle. All the pedals are spinning at the same rate, but each rider can choose to apply a certain amount of torque (for a bicycle, this manifests as each crank deflecting slightly vs. the natural position; for a generator, this ends up being a phase difference between the low vs. high load states). If you can coordinate everyone together, then you can slowly increase the speed as they pedal harder, or the reverse.