Stability of grids with a high percentage of asynchronous sourcing

[Someone]

No-one is trying to blindly synchronise generation without looking at the local phase.

Yes, but that´s exactly what would be required if a grid should break down by one failing oscillator.

You may need to restate that?

Hard to believe a grid could (or should) fail from a single oscillator failure. A single generator might go offline, but that should not be a problem for the wider grid.

People keep trying to say there is some required [long term] integral accuracy for a generators local time reference. That is untrue. Generators currently need:
a) a somewhat accurate frequency reference (which is easily provided by low cost crystal oscillators) to participate in maintaining frequency
b) some way to align their rotation (synthetic or not) with the local phase before connecting power

Billing might care if it's 2 cycles past midnight, but a generator doesnt care about absolute time.

[nctnico]

True. But imagina a long power line (say 300km) with two identical generators at each end. The impedance of the power line will be high enough that the generators can run almost at will. Ofcourse you'd start with one generator and the other one can be started in sync (to match the phase), but once running at full power the synchronisation effect won't be very strong. And which one is the leader?

A similar effect can be observed with classic ethernet or long RS485 busses. Two stations far apart can start transmitting without detecting the other one is garbling the messages due to the impedance of the transmission line.

See this presentation: https://wsts.atis.org/wp-content/uploads/2021/03/Challenges_to_Updating_Timing_In_The_Power_Grid.Jones-Shoenwald-Tuffner.pdf

And NIST also published a report on time synchronisation needed to monitor the grid stability: https://nvlpubs.nist.gov/nistpubs/specialpublications/nist.sp.1500-08.pdf

[Phil1977]

True. But imagina a long power line (say 300km) with two identical generators at each end. The impedance of the power line will be high enough that the generators can run almost at will. Ofcourse you'd start with one generator and the other one can be started in sync (to match the phase), but once running at full power the synchronisation effect won't be very strong. And which one is the leader?

If the line impedance would be that high you couldn't operate a power grid at all, this would mean that not much power is transferred at all.

If both generators have a good coupling by a low-impedant grid (as a grid needs to be) they keep themself in sync. You could do the maths, or to make it easier you can imagine that even parts of a degree of phase mismatch would induce a high reactive power over the grid line. This reactive power slows one generator down and speeds the other one up - up to the point when they are in sync again.

You can reproduce exactly the same electrodynamic behaviour with electronic inverters. Mechanic assemblies with large inertia make it better descriptive, but the principle is exactly the same.

And NIST also published a report on time synchronisation needed to monitor the grid stability: https://nvlpubs.nist.gov/nistpubs/specialpublications/nist.sp.1500-08.pdf

Okay, now I see the problem: We we´re talking in this thread about power generation. NIST is talking in the paper about power distribution.

If you want to do power distribution with a large scale grid that has to cope with phase mismatch by lines of different length and impedance and homogenizing the flow of power over the available lines, then things are infinitely more complicated. But there are very smart people (e.g. at NIST) who will take this challenge.

But for the sourcing itself - keep things easy!

[nctnico]

In the end grid based power generation and distribution go hand in hand. Potato potaato

[Phil1977]

In the end grid based power generation and distribution go hand in hand. Potato potaato

Generation, distribution and consumption should even go hand in hand 

No one says precise timing is not important - but it´s no special problem of de-central generation.

[woodchips]

Rather than try and remember stuff from years ago, I have trawled through many relevant books. These are all 50-100 years old.

There is no central reference voltage or frequency on the grid. It is distributed amongst all the generators. In the UK typical consumption now is about 30GW, so even the largest 1GW generator is very small, so no individual generator can vary either voltage or frequency. This is what is called an infinite busbar.

All power stations read the current busbar voltage and frequency. If any change is needed then many have to work together.

The only way to alter voltage or frequency is to increase or decrease the steam going to the turbines, many or them. The system inertia is such that a change is only possible after many minutes or hours.

I assume that frequency is measured by cycle counting, say for 10 minutes it counts every cycle and stops, it should be 30,000 so easy to decide whether you need one or two cycles up or down. And that is the stability that is looked for, all the synchonous clocks need to keep time.

Now that renewables have arrived, then all the photos I have seen of wind turbines have a rotating alternator/induction generator so they aid the frequency stability of the grid. An inverter is a different thing altogether, there is no inertia in it at all. So it must sample the connected grid to obtain frequency and voltage, then equal it. This does not work if all the rotating generators suddenly disappear.

This is more complex than it seems, an overhead line is limited to about 100 miles but even that will have a noticeable phase shift, only 1 degree will result in lots of losses and heat. So if you have inverter farms that far apart then how do they synchronise, atomic clocks seem useless, they are not accurate to the line state at each end.

Is this one of the new fangled whizzy improvements, rotating metal to inverters, that aren't quite so red hot, literally? Bit like mass software updates?

It is interesting though, keep the comments coming please.

[Siwastaja]

An inverter is a different thing altogether, there is no inertia in it at all. So it must sample the connected grid to obtain frequency and voltage, then equal it. This does not work if all the rotating generators suddenly disappear.

It's completely beyond me why people keep repeating these lies, like parrots, despite these have been shown complete lies dozens and dozens of times, in every thread Please just stop.

Given that inverters have already implemented these inertia-emulating curves for many years (another question is, is that all currently enabled, because there is still little need), this is really getting into "look at this flat earth" territory.

This is more complex than it seems, an overhead line is limited to about 100 miles but even that will have a noticeable phase shift, only 1 degree will result in lots of losses and heat. So if you have inverter farms that far apart then how do they synchronise, atomic clocks seem useless, they are not accurate to the line state at each end.

This idea of somehow having to synchronize inverters into operating in lockstep and compensating some power transfer delay is absolutely ridiculous - obviously other types of generators never did this either. Because these idiots parrotting it never stop regardless of how many times they have been shown being wrong, I'm now more interested about where is this idea originally coming from? It's weird enough I don't believe so many people have come up with it independently.

In other words, I'm saying this is probably some kind of semi-organized propaganda. I might be wrong though.

[Marco]

Commercial system, not research.

https://www.fimer.com/sites/default/files/FIMER_Brochure-Utility%20Applications_EN_revB_1.pdf

"Full four quadrant active power and reactive power support"

It's not a complete multi-functional inverter, but the systems can already provide power support (ie. sink a bit during the cycle). If the electricity companies want more, they should just demand it from utility scale solar. Technology is ready.

[Geoff-AU]

This is more complex than it seems, an overhead line is limited to about 100 miles but even that will have a noticeable phase shift, only 1 degree will result in lots of losses and heat.

The wavelength of 50Hz at the speed of light is 6,000km.  1 degree of phase shift is 16km.  So if 16km is enough phase shift to cause problems, how does anything work at all?

100 miles, or 160km, would be 10 degrees of phase shift.  That sounds quite catastrophic.  Except... it isn't.  There are many overhead lines that span substantially more than 160km.

I suggest there is a logical fallacy in your reasoning somewhere. 

[zilp]

An inverter is a different thing altogether, there is no inertia in it at all. So it must sample the connected grid to obtain frequency and voltage, then equal it. This does not work if all the rotating generators suddenly disappear.

It's completely beyond me why people keep repeating these lies, like parrots, despite these have been shown complete lies dozens and dozens of times, in every thread Please just stop.

In particular: It has been explained in this very thread, repeatedly.

And not only that, it is also just obvious that it is nonsense. Like, how does anyone who has any idea how an inverter works even get the idea that it somehow could be impossible for an inverter to change output power based on grid frequency fluctuations?!

But, yeah, I agree that it is the perfect propaganda lie: People interested in technology might have heard that the inertia of rotating masses in thermal power plants stabilize the grid, and obviously, your typical solid state inverter doesn't have any rotating mass ... so, that's probably an effective hook to make people who lack any deeper understanding believe that therefore, inverters can't stabilize the grid, and that they are smart enough to understand this while everyone else is blind to this truth because they don't understand technology, which immunizes them against criticism against their position, because that criticism then must be motivated by the "green ideology" of people who don't understand technology ...

[Siwastaja]

It's not a complete multi-functional inverter, but the systems can already provide power support (ie. sink a bit during the cycle). If the electricity companies want more, they should just demand it from utility scale solar. Technology is ready.

This has even been a legal requirement for a long time, IIRC from 2010 or so in Germany for example. More than 10 years!

Every cheap Chinese inverter (e.g. Solis datasheet in attachment) can participate, either by being configured with a fixed phase shift; or, if at some point it seems necessary to add centralized control for a swarm of smaller inverters acting towards the same goal (which could be useful for long-term frequency stabilization, the thing for those retro clocks), then small IoT dongles with MODBUS/RS485 can be simply added. In fact most of these inverters are already in manufacturer's cloud service (because installers can that way monitor and debug).

And to be clear, every generator - inverter or rotating - has to synchronize to the grid (just by sensing the local grid voltage, not by some weird timing-critical communication scheme), and only have two parameters to control, both slowly varying (within seconds): real power output, and slight adjustment to reactive power (lagging or leading power factor). That's it. Traditionally these parameters were controlled by actual people in control rooms based on some very simple rules of thumb. This is all trivially easy to algorithmically replicate and this has been done decades ago.

[Someone]

An inverter is a different thing altogether, there is no inertia in it at all. So it must sample the connected grid to obtain frequency and voltage, then equal it. This does not work if all the rotating generators suddenly disappear.

It's completely beyond me why people keep repeating these lies, like parrots, despite these have been shown complete lies dozens and dozens of times, in every thread Please just stop.

It doesnt help when the operators simplify it down to that "doesnt work" claim:
https://www.nationalgrideso.com/electricity-explained/how-do-we-balance-grid/what-inertia
Which is untrue, out of date, yet convenient to protect entrenched investments/decisions.
Yet in other publications the same organisation is proposing? currently? using inverters for the inertia:
https://globalpst.org/wp-content/uploads/ESO-inertia-management.pdf
https://www.nationalgrideso.com/document/278171/download
"GFC"/"GFM"/"VSM" all having been demonstrated and known to be practical, by ESO.

[.RC.]

All I know is I expect down the track with millions of generators all connected together all relying on third party support there will be a huge catastrophe as there will be an event that will wipe a lot of these micro generators out, parts will be impossible to come by to repair.

The KISS theory has well and truly been thrown out the window by this rush into wind, solar and what not.

[zilp]

All I know is I expect down the track with millions of generators all connected together all relying on third party support there will be a huge catastrophe as there will be an event that will wipe a lot of these micro generators out, parts will be impossible to come by to repair.

The KISS theory has well and truly been thrown out the window by this rush into wind, solar and what not.

All I know is I expect down the track with millions of generators all connected together all providing each other support there will never again be a huge catastrophe as there will never be an event that will wipe a lot of these micro generators out, so availability of spares doesn't matter much.

Luckily, I can just make stuff up to cancel out the stuff that you make up.

Now, can we get back to falsifiable theories of grid operation rather than ideologically informed gut feelings?

[Siwastaja]

all relying on third party support

Spotted the exact lie injected here. In reality, obviously solar inverters work independently, without any support, communication, updates or whatever.

[.RC.]

Actually I meant in the case of Australia absolutely everything is made overseas, so we are reliant on third party countries, some of which hold expansion plans.

I do not think we could make a SMD resistor here. Or maybe one company in Sydney can.

We certainly could not make an inverter from local parts.  These "big batteries".   Someone could go out with a gun with silencer and subsonic ammunition and easily take out an entire storage facility without anyone any the wiser.

[Geoff-AU]

we are reliant on third party

You're reliant on so many 3rd parties you take for granted already.  Can any of us survive completely independently of others?  I doubt it.  I wouldn't have the first idea how to feed myself from first principles and I hope I never have to try.

[woodchips]

Some more comments. Why can't someone describe, in words of one syllable, the process by which inverter number 53 syncs itself to the other 52 on a local grid. The abuse really doesn't get anyone any further, I have spent lots of time reading books about the start of the grid and how the people then tried and tested to prove that, yes, you could connect synchronous generators in parallel, but not series.

What it the normal size of inverters? If they are ones off solar panels then 10kW seems to be it, if a reverse car battery charger then 100kW is possible. These are peanuts compared to almost any rotating generator. Doesn't this huge reduction in the size of each individual energy source cause problems?

Even with modern inverters, if you get your voltage to be different to the grid voltage then current will flow, somewhere. So, again, in words of one syllable why not explain how the tiny little inverters cope with this voltage difference. Of course in a rotating grid there are just two controlling adjustments possible, steam and excitation, what replaces these?

[Siwastaja]

the process by which inverter number 53 syncs itself to the other 52 on a local grid.

It's simple: it doesn't. It only syncs to the voltage it measures right where it is connected.

It can be as simple as measuring voltage, and controlling current controller's setpoint to be in direct relation with measured voltage.

It can work as simply as this:
* Measure instantaneous line voltage (e.g. 100V)
* Multiple this by a constant, say 0.1, to decide injected current. e.g. 100 * 0.1 = 10 (A)
* Set current controller loop to output 10A
* Current controller measures current through a shunt resistor and sees less than 10A going
* A transistor is turned on, and current through the inductor in the converter starts to rise
* Once 10A is flowing, transistor is turned off

Repeat this process thousands of times a second, and you have an inverter which "syncs to grid" and outputs power with PF1.0.

The above constant, 0.1 in this example, needs to be adjusted in a slowed control loop so that proper PV DC input voltage is maintained.

This is all electronic equivalent to what rotating machines do. They also sync themselves to the grid locally.

if you get your voltage to be different to the grid voltage then current will flow, somewhere.

By Ohm's law, yes. This small voltage difference is what "drives" current, in all power distribution systems, AC or DC. Or you can swap the thinking around and say that driving current into lines causes a small voltage drop; the same thing. It is just how it works. It is interesting how you find somehow the fact that there is current flowing in power grid somehow weird or problematic to mention it like this. I think to many, even complete laymen, it is obvious that current is flowing in power lines (grid or their home wiring).

[gf]

Here is an overview of control methods for grid forming inverters: https://www.mdpi.com/1996-1073/13/10/2589

Conventional small grid-tied inverters are, however, not grid forming, but their control loop rather makes them behave like an AC current source, where the injected current is phase-locked to the grid voltage (possibly with some phase offset, if the inverter is supposed to supply reactive power).

[zilp]

Some more comments. Why can't someone describe, in words of one syllable, the process by which inverter number 53 syncs itself to the other 52 on a local grid.

It does so by means of a PLL, usually implemented in software.

It's not that you are wrong when you say that the inverter "samples the grid to obtain frequency and voltage". Your error is in assuming that a turbine/generator machine doesn't do the same. When you bring up a rotating generator, you also regulate its rotation rate until it is in sync with the grid, then you connect it. It's no different with an inverter. The only difference is that the inverter has basically no limit on how fast it can adjust frequency and phase, so an inverter can in principle achieve grid sync in a few milliseconds, which a rotating generator can't do.

What it the normal size of inverters? If they are ones off solar panels then 10kW seems to be it, if a reverse car battery charger then 100kW is possible. These are peanuts compared to almost any rotating generator. Doesn't this huge reduction in the size of each individual energy source cause problems?

What problems?

Even with modern inverters, if you get your voltage to be different to the grid voltage then current will flow, somewhere. So, again, in words of one syllable why not explain how the tiny little inverters cope with this voltage difference. Of course in a rotating grid there are just two controlling adjustments possible, steam and excitation, what replaces these?

They don't "cope" with anything, the creation of that voltage difference is their job, if they didn't create a voltage difference, they wouldn't work.

I'm not sure what your idea is of how an inverter works, but an inverter certainly doesn't have some kind of fixed output voltage that then can somehow be massively mismatched with the grid, if that is what you are thinking!?

The inverter output stage has a power target (that is determined, in the case of solar inverters, by the MPPT of the input stage) and will thus create sufficient voltage to push that much power into the grid. How much voltage that is depends on the voltage and impedance of the local grid connection and will be regulated cycle by cycle. Necessarily, the voltage needs to be higher than the "idle" voltage of the connection, or else no power would flow into the grid. The inverter will simply continuously measure voltage and current and will adjust the voltage such that the product of voltage and current matches the power target. At the same time, it will monitor the voltage so as to not exceed certain limits, and will reduce or completely disable output as needed.

There isn't really any "replacement for steam and excitation", simply because these aren't mechanical systems. This is software running on a microprocessor that switches power transistors on and off in order to achieve the desired voltage and current wave forms. Also, having to regulate steam and excitation is not somehow a great feature of thermal power plants, it's rather a limitation, so it doesn't make a whole lot of sense to want to replace them. What you are actually trying to regulate is the voltage and frequency/phase of the electrical output. With thermal power plants, you only can do that the roundabout way of changing steam and excitation. With an inverter, you can just change them instantaneously if you write the software that way. Or you can change them slowly, if you write the software that way. An inverter can simply change power output rapidly, cycle to cycle, something that no thermal power plant can do, and it is all up to the software how you regulate power output.

[Someone]

What it the normal size of inverters? If they are ones off solar panels then 10kW seems to be it, if a reverse car battery charger then 100kW is possible. These are peanuts compared to almost any rotating generator. Doesn't this huge reduction in the size of each individual energy source cause problems?

The onus is on you to explain why a collection of smaller generators causes problems. Asking others to find justification for your unsupported ideas just points to your inability to even enter the conversation.

Even with modern inverters, if you get your voltage to be different to the grid voltage then current will flow, somewhere. So, again, in words of one syllable why not explain how the tiny little inverters cope with this voltage difference.

How is this any different to a grid with numerous rotating mass generators? You have chosen to say there is some distinction, but have not pointed to any.

Of course in a rotating grid there are just two controlling adjustments possible, steam and excitation, what replaces these?

The grid as a system has no concept of those, it cannot be reduced to "just two" things and certainly not to those things.

[Someone]

Here is an overview of control methods for grid forming inverters: https://www.mdpi.com/1996-1073/13/10/2589

Conventional small grid-tied inverters are, however, not grid forming, but their control loop rather makes them behave like an AC current source, where the injected current is phase-locked to the grid voltage (possibly with some phase offset, if the inverter is supposed to supply reactive power).

This is part of the subtle differences that are easy to produce misleading comments from.

That existing inverters (generally) have not provided frequency inertia does not imply that inverters alone cannot provide inertia to fully support the grid. A grid without any rotating mass is known to be practical with today's technology, and has already been demonstrated. For most inverters thats just a change of control parameters or a software update away, the physical capability is already out there.

I'll strongly disagree with that linked papers binary distinction of gird forming [voltage controlled] vs current controlled, as current controlled inverters are equally capable of producing frequency inertia with suitable PLL tuning (which is/was explicitly prohibited by grid operators).

[Phil1977]

Even if for some hypothetical reason a grid would need rotating mass you could install so called "synchronous condensors" which principally are just a large electric motor / generator connected to the grid. In the grid or in industrial applications they are used to reduce reactive power end enhance grid transmission efficiency:

https://en.wikipedia.org/wiki/Synchronous_condenser

This device has a large rotating mass that acts as a kind of flywheel with a certain level of rotational energy. If voltage or current in the net are out of phase, or if load spikes need to be dampened then energy is fed into or taken from this flywheel.

It´s really not difficult to setup a good physical model when electricity needs to be converted into rotation or when it´s necessary to generate electricity from rotation. You can take exactly these rules to build a circuit that stores or takes energy from a big capacitor.

This is - extremely simplified - what's done in grid stabilizing inverters. You have a well buffered DC bus that can feed the grid with power exactly when a beautiful sine wave needs it - or even take momentary power in certain circumstances.

You don't need an atomic clock or anything very exotic for this. The capacitor you need in the inverters isn't even that big, not much more than you anyhow need for the conversion circuit.

It´s sad but true that many people nowadays are afraid of the future and think they can keep anything as it is by denying progress. They just say progress is impossible because they are afraid of it. And to justify this some people shamelessly lie and other frightened people spread these lies. This grid stability thing is one of these lies.

[nctnico]

This device has a large rotating mass that acts as a kind of flywheel with a certain level of rotational energy. If voltage or current in the net are out of phase, or if load spikes need to be dampened then energy is fed into or taken from this flywheel.

It´s really not difficult to setup a good physical model when electricity needs to be converted into rotation or when it´s necessary to generate electricity from rotation. You can take exactly these rules to build a circuit that stores or takes energy from a big capacitor.

This is - extremely simplified - what's done in grid stabilizing inverters. You have a well buffered DC bus that can feed the grid with power exactly when a beautiful sine wave needs it - or even take momentary power in certain circumstances.

But this is what I've been claiming all along: you will need devices to govern the grid stability regardless how they are physically constructed. Decentralised generation or not, there have to be devices which govern synchronisity. Like an orchestra needs a director. But those grid synchronising devices will need to be loud enough to overcome the noise as each micro inverter feeding the grid adds noise due to component tolerances and measuring errors.

And for sure you don't need an atomic clock perse, but it certainly is a nice to have as even today clocks depend on the grid frequency being stable most of the time.