Grid Stability and Renewable Power

No. Grid storage power storage provides exactly that. Pumped hydroelectric power is the traditional scheme but grid-scale batteries are faster.

Australian rooftop solar installation are starting to include enough battery storage to keep the house running overnight. The local electric utilities aren't enthusiastic about buying surplus electric power from rooftop installations during the day, so it's more profitable for the house-holder to spend of the order of $1,000 on a battery and use up the excess power themselves.

If you have grid-scale batteries the frequency doesn't change much at all - they can provide cycle-to-cycle phase correction.

Half the capacity of South Australia's famous Tesla grid-scale batter is devoted to doing just that, and it paid off the price of the whole battery within two years.

Not really. The sun goes down at night. Wind turbines are even less predictable.

It's not what they are doing in South Australia

It's a sub-optimal solution.

The South Australian story makes it obvious that this is nonsense.

But not at the right frequency. Inverters are ubiquitous.

Older systems with high percentage of old style synchronized high inertia generators provided good stability

In the future more generated power will be decentralized, so you need a distributed communication channel to set the duty point of the majority of the power generated

AFAIK most solar systems are current sources with no knowledge of grid state, resulting in VAR squishing around on the grid

The people who turn that low voltage direct current output into mains frequency input into the grid are responsible for getting the phase right.

If the utilities don't manage them well enough to prevent "VAR squishing around the grid" that's simple incompetence.

Ricky wrote: ==========

** They do that right now and for the past 100 years.

Hydro power stations are the main frequency regulating elements in a grid. Certainly here on the east cost of Australia ( Snowy River Scheme) and I bet in any other place blessed with hydro.

Reasons being that generation can be fine controlled almost instantly in response to frequency deviations and permanently availability when a number of dams and stations are involved.

The permanently part arises form the ability pump water up hill and keep a few dams full at all times. Sun light and wind are God's domain, not human controlled.

...... Phil

What is the 'inertia' analogy, exactly? Most renewable power sources aren't like hydroelectric, with large stored reserves and quick access to extra generating capacity, but a 'grid' can have (and use) a variety of sources, including some with stored reserves, in cooperative fashion.

It's a stable grid if its management has the right combination of feedback and time delays... just like a compensated op amp.

[about solar photovoltaics]

If by that they mean inverters that aren't managed as part of the overall grid, but which are always-full-power, then the feedback requirement isn't met. So it does make some sense. The photovoltaic output would have to exceed the sum of easily-controlled other sources, or have abrupt transitions that exceed slew rate of the grid-control options. Or, the grid control would have to be badly designed.

On a sunny day (Mon, 18 Apr 2022 00:51:00 -0700 (PDT)) it happened whit3rd snipped-for-privacy@gmail.com wrote in snipped-for-privacy@googlegroups.com:

Molten salt sun power plants use 'inertia' to supply power day and night:

Part of the problem in the UK was that because of micro generation on individual home roofs have to self protect their 4kW rated kit if the load gets too far off specification they each make an independent decision to disconnect leading to a runaway cascade failure.

That is part of what the technical investigation into the big UK powercut that took down most of London and the SE in August 2019.

The other snag was that the low frequency demand disconnect system dropped both load and active generation capacity on home roofs so that the numbers no longer added up. Net load shed was much less than the absolute load shed (because of local PV generation on roofs).

It is a case study in how adding renewables to the mix effectively destabilised the network because some of the new protections on big offshore windfarms were untested/incorrect and conceptually flawed.

The grid is generally balanced by dumping all residual power into electrolytic aluminium or brine plants that can absorb any amount of energy and can change how much they take in an instant. They do require a certain amount of power to stay hot/warm but can vary their consumption by two or possibly three orders of magnitude when required.

No. The individual domestic systems will each try to save themselves when the network conditions become adverse. Only really big generators can provide the inertia (pumped storage or solid state huge batteries)

Neither can provide a long term solution so if some more conventional power doesn't come back onstream before it runs out you are stuck.

The ultimate sanction is that the grid will shed great chunks of load until it is able to get the frequency back under control. The calculation of how much has been greatly complicated by solar PV.

There is no point in operating at anything other than peak efficiency. The grid is always balanced for consumption and generation with the loads of last resort taking up any instantaneous slack. They also get dumped off first if there is a glitch.

How? Each of the toy systems on a home will make its own decision about when to drop off as the frequency goes out of spec. You probably could allow them some more leeway to stay on grid for longer.

Nothing can really get around the fact that if the house they are on gets disconnected from the grid by load shedding their contribution (which during daytime might well be net positive) is lost.

The calculation that wasn't allowed for in the UK is that with domestic generation on home roofs and on a sunny day when you shed "load" you will also shed a whole bunch of local solar PV generation as well.

This is an active area of R&D, and there have been some pretty significant developments which seem to have gone by unnoticed by most, namely code changes mandating the use of inverters compliant with UL1741 SA. This changed the requirements for grid connected inverters from a very simple "disconnect immediately when the grid goes out of a very narrow definition of normal and reconnect after the grid has been normal for 5 minutes" requirement ("Grid Interactive"), to a much more sophisticated "Grid Support" requirement, where inverters are required to help stabilize the grid. They can do this more effectively than large rotating generators because their response time is /much/ faster.

UL1741 SA (Supplement A) has since been rolled into Rev 1 of the base UL1741 spec, but SA is still a good search term. The only accurate description of these requirements I have found, short of shelling out thousands on specs, is a one hour seminar available on the UL web site, registration required. There is a lot of BS on all other sources I have seen, mostly people don't seem to understand the relationship between reactive power and grid voltage (Grid Support inverters supply reactive power to the grid to help correct under voltage even if they are already at maximum output; they can deliver significant reactive power with only a small reduction in real power, and reactive power is more effective in boosting grid voltage than real power due to the characteristics of the rotating generators and motors on the grid. (By convention capacitors supply reactive power and inductive loads use it.)

Requirements for grid connection of any power source are published in a utilities SRD (Source Requirements Document), and as far as I know all US utilities SRDs have required Grid Support inverters since 2020, and they are also required by the current NEC.

Some utilities, Hawaii and possibly others, also reserve the right to require a SCADA-like monitoring and control network connection to your inverter - they want your knobs :-).

There have been similar code changes in Canada and the EU.

IEEE refers to grid connected inverters as "static synchronous generators" and the old-fashioned type as "rotary synchronous generators" in newer specs, liberally referenced by UL.

Glen

OTHER than hydro... of course.

Let's leave the small, domestic systems out of the conversation. The particular point someone was making was that no inverters used with wind power (or solar farms) has the ability to help stabilize the grid, because there is no rotating inertia. It was not claimed that this was not possible, but it was implied by pointing out no one had done this yet and it would be a very useful feature.

Seems to me it would require some way of increasing the power output, which means the facility has to run below optimal efficiency to have anything in reserve.

The advantage of natural inertia, is the continuous nature. As much energy as is needed is available if you are able to tolerate the reduction in frequency. Of course, there is a limit to the inertia available, but it seems to do the job pretty well in most cases, while currently we seem to get nothing from solar and wind power facilities.

Ricky wrote: ===========

** What happens now is that inverters feeding the grid *track* the existing frequency - cos they are minor players in supplying the load. But what if that were not the case, they became the majority suppliers and and instead were locked to a central clock ?

Rotating machines would then follow them.

...... Phil

That's a weird thing to say. If the "inverters" were locked to a central clock and driving the grid, rotating generators would be no different than they are now, following the grid.

The grid is very complex. Many different actors, who is leading and who is following in this dance? It's not quite so simple. That's why rotational inertia is useful.

Ricky the IDIOT wrote: ==================

** Whaaaaaaatttttt ??????

Insane, retarded crap.

...... Phil

Synchronizing grid tied inverters and generators using GPS clocks and GPS disciplined oscillators is a common topic for green research papers. For example: "How Microsynchrophasors Could Keep Solar-Saturated Grids Stable"

The fun begins if the grid frequency slows down a little due to a decrease in source supply or an increase in load. A short while later, the generators are adjusted to bring everything back to exactly

50 or 60Hz. However, that's not good because the frequency also needs to be adjusted to compensate for the time lost during the power sag. Otherwise, all the power line driven synchronous motor clocks would runs slow. So, the frequency of the entire grid needs to be increased slightly until the lost milliseconds are recovered, when it is now safe to return to exactly 50 or 60Hz. This explains the basics of how it's done: Notice that the example shows that Swiss time was 160 seconds behind UTC in June 2013. At 50Hz, that's 160sec * 50cycles/sec = 8,000 clock cycles that need to added to the grid for grid time to catch up with UTC time. Looking at the graph (blue line), time still hasn't caught up 6 months later.

Bottom line is that synchronizing grid tied inverters is certainly possible, but isn't quite as simple as it might initially appear.

Trivia For UK grid:

snipped-for-privacy@gmail.com wrote: ====================== Phil Allison

** Nice to know, I was just speculating. ** Errrr - why ??

Alternators naturally slow in reaction to load, but not inverters. Plus all alternators in a grid are locked together in phase.

...... Phil

Ah, Phil speak for he doesn't understand.

I don't know about Switzerland, but in the US, didn't they throw in the towel on supporting synchronous clocks? I think that happened over 10 years ago. Maybe I was misinformed. I found an article at the IEEE from 2011 about a year long experiment where they were going to stop correcting the grid to see how many people complained. I didn't find anything about the result.

This is probably the 2011 paper about the test:

Reports started to appear a few years later. However, I can't tell if the 1 year test was actually performed. This report looks like it was done using historical data from a power line frequency monitoring network: "Impacts of Power Grid Frequency Deviation on Time Error of Synchronous Electric Clock and Worldwide Power System Practices on Time Error Correction"

"On the other hand, the identification results present that up to the end of 2016, many electric utilities around the world, especially in North America and Europe, provided the TEC service to periodically remove the accumulative time error of synchronous electric clocks."

This CAISO document indicates that TEC (time error correction) was active in western USA at least up to 2019:

This 2021 document indicates that in the event of an emergency, time error correction can be temporarily suspended until things sort themselves out:

Kinda looks like TEC is currently alive and well, at least for some grid operators.

Drivel: Old technology doesn't completely die out, even after several generations of superior technologies. For example:

Time error correction is no longer done in the US after 2017:

"... so last year, the correction part was quietly eliminated by the Federal Energy Regulatory Commission."