@transparent my reading of the report P28 (and probably limited understanding) is there were 3 oscillation events 11am, 12 and then the catastrophic one @12:30 and after all 3 the demand on the system increased and the demand increase is possibly due the behind the meter PV cutting out. The system recovered after the first two oscillations. The failure cascade truly started from the additional 355MW generation loss due to transformer trip in the substation in the Granada area. (P103 listing) following the 317MW of probably private PV drop-out causing a load increase.
concerning certainly that such voltage and frequency oscillations are not uncommon.
2kW + Growatt & 4kW +Sunnyboy PV on south-facing roof Solar thermal. 9.5kWh Givenergy battery with AC3. MVHR. Vaillant 7kW ASHP (very pleased with it) open system operating on WC
Apologies @judith I could've made my point clearer if I hadn't clipped my screenshot that small.
Let me post the relevant section of the Report again, but see my highlights in the previous paragraph, which was absent in my earlier post
The Committee have differentiated between the two earlier oscillations and the Incident itself.
I'm sticking with their description... ... and I have another reason for doing so.
All of the other oscillations and trips could be attributed to mechanical failures or deliberate actions by humans (benign or malicious).
But the loss of 317MW of embedded generation is quite different. It's simply a cascade effect caused by the solar-inverters complying with their [G98] regulations.
GB is in the same situation. System resilience is being diminished because we have (digital) inverters supplying power at all voltage levels of the grid.
@transparent I’ve understood your point now. The point I was trying to make (probably not clearly) is the fairly fast (<1Hz) voltage oscillations tripped the domestic PV but the system recovered but the whole system went down later when other equipment also reacted to oscillations. We don’t know enough yet to compare easily to the UK system but there are common concerns.
The UK domestic roof-top PV is probably never going to be up to 1% of total load but in a local area it might be closer to 1%. We’re fortunate here that we’ve not had a local high voltage generation constraint on our domestic PV but as you posted above they do happen.
It gives the grid more reasons to suppress oscillation since the side effects of equipment tripping can be a problem to handle.
What I don’t understand is what stopped the earlier oscillations since the correction happened after they stopped ‘naturally’. And the 3rd and final was different it seems.
2kW + Growatt & 4kW +Sunnyboy PV on south-facing roof Solar thermal. 9.5kWh Givenergy battery with AC3. MVHR. Vaillant 7kW ASHP (very pleased with it) open system operating on WC
The UK domestic roof-top PV is probably never going to be up to 1% of total load but in a local area it might be closer to 1%.
It rather depends on what we call 'local area' and 'total load'.
Let's re-phrase 'total load' as 'power being handled by the transformer'. That avoids the issue of which way the power is going!
Here's a timeline from a local substation at housing estate in Wales:
As day breaks there was fog or cloud cover, which disperses around 10:00.
The houses connected to the phase coloured turquoise have more solar PV than those on the other two phases. Consequently, the turquoise houses now make much less demand on the grid, and the current through the transformer falls.
Current returning through the neutral rises correspondingly... ... to such an alarming level that it's greater than current through the phase wires at the points I've indicated with red lines.
Excess solar generation is extremely common in the West Country, where I live.
That from domestic production obviously cannot be curtailed dynamically by the DNO. To keep the grid operating within statutory limits, they must deploy Active Network Management (ANM) which 'rejects' export from commercial sites in the area.
I have a few of those ANM plots, produced as forecasts to companies seeking grid connections. It's very easy to see from those plots the curtailment is to be implemented due to excess solar:
The quantity of rooftop solar being exported to the West Country distribution grid is far above 1% however you define it.
Oscillations are a 'normal' part of grid operation.
There's no central 50Hz master-clock.
Every generation/storage site is basically synchronising itself to all the others, which is what we 'see' as oscillation. The speed at which a generator synchronises is important, which is another reason why Rate of Change of Frequency is of greater significance than the actual frequency.
Attempting to synchronise too slowly means that the grid can go down before your generator can help to stabilise it. Synchronising faster can help prevent sudden fluctuations when faults occur, but it risks causing wild fluctuations too.
What you want is a whole bunch of rotating turbines which react at a predictable speed.
The more generation that gets supplied by (electronic) inverters, the more likely it is that oscillations will happen wildly.
@transparent thanks I’ve learned a lot from your post above. I’m intrigued that you are suggesting potentially malicious actors? We’ve all got a lot to learn.
2kW + Growatt & 4kW +Sunnyboy PV on south-facing roof Solar thermal. 9.5kWh Givenergy battery with AC3. MVHR. Vaillant 7kW ASHP (very pleased with it) open system operating on WC
It was a story carried by three UK-based media outlets on the following weekend, but they misidentified the threat as an embedded "Kill Switch". See this article by LBC.
The original Reuters announcement is derived from a leaked report in USA.
American engineers disassembled inverters which were connecting commercial solar arrays to the grid. What they have found are six cases where there are additional remote-control communications pathways, not referred to in the manufacturers’ specifications. It has subsequently been reported that these secondary paths are cellular in nature, and that the investigations in USA have been ongoing for over a year.
Such an approach can facilitate the inverters being controlled by a remote 3rd-party, in addition to the US company which is authorised to export power to the grid.
We need to consider whether unauthorised communications and control mechanisms have been used solely to target the USA. Or is it equally likely that such pathways are routinely present on inverters designed and manufactured in the Far East, but the US engineers are the only ones to have discovered them?
The situation with a storage battery would be more serious if they too had hidden remote control within inverters. It's easier to crash the electricity grid by suddenly charging a BESS. That takes energy away from the supply system, and lowers the voltage. Other grid-tied inverters will respond to that low-voltage by disconnecting their own grid export.
That's what I refer to as a Cascade Event. You don't need many compromised inverters to start such a chain reaction.
This post was modified 3 weeks ago 2 times by Transparent
I've checked the Spanish equivalent to our G98 certification regarding export to the grid. DNOs in Britain usually permit export at 16A (per phase) for inverters with G98 approval from the ENA.
In Spain, Royal Decree 244/2019 limits grid export to ZERO by default. Technical Standard UNE 217001 defines how an inverter must comply.
Premises with PV panels are permitted to generate for Self-Consumption only, which may optionally be stored in batteries.
Collective Self Consumption is an extension which permits multiple premises to share solar-generation, but export to the wider grid must still be zero.
The default household supply is just 3.5kW (16A, single-phase 220v), limited by a re-settable trip called an IPC.
You can pay more to obtain a larger grid supply, up to 45A. Above that you will be offered a 3-phase supply.
Thus if your household is using self-consumption above the rating of your IPC, and cloud obscures the sun, your trip will 'blow'.
That provides a crude protection against surge demand in an area.
If Zero Export is the norm, then I don't (yet) understand why the latest ENTSO-e report suggests the possibility of small embedded (solar) inverters becoming disconnected from the grid.
[... ]there was a loss of 208 MW identified distributed wind and solar generators in northern and southern Spain, as well as an increase in net load in the distribution grids of approximately 317 MW, which might be due to the disconnection of small embedded generators < 1 MW (mainly rooftop PV) or to an actual increase in load or to a combination of both.
The initial report of the investigation of ENTSOE EU was published
The report says that this is being investigated by the RCC investigation (Regional Control Centre). This RCC handles France, Spain and Portugal. Coreso is the organization owned by these and a few other European network operators including Belgium, UK, Germany, Italy and Ireland.
It is the first time that a cascading series of disconnections of generation components along with voltage increases has been part of the sequence of events leading to a blackout in the Continental Europe Synchronous Area.
The expert panel will assess if the most relevant requirements for this incident arising from European and/or national regulations and contracts – such as those in the SO GL, the Connection Network Codes (e.g. RfG and DCC), the Emergency and Restoration Network Code (ER NC), Synchronous Area Framework Agreement (SAFA), national regulations (e.g. voltage control and voltage operational limits) or the TSO-DSO protocol on voltage control – were fulfilled.
The Expert Panel envisages completing the final report within approximately four months following the publication of this factual report. However, this timeline is purely indicative.
So call me an utter geek [🤓] not only was I hanging out with transmission system operators in Brussels but I also spent time in England with Elexon training on the.... (wait for it...) Balancing Mechanism.
Holy smoke it's complicated but it's not possible to understand the technicals of our wondrous grid without it.
I've got more to say about this (but not on a Saturday night 😁) because it's a key part of the ENTSO report & also the public version of the Spanish government report.
which didn't match the bad news which certain media outlets were looking for:
Do these guys not even understand the basic science of what 6GW extra supply looks like?
They don't want or need to understand, they have an agenda to promote thanks to their political angle and fossil-fuel backers. I have a big fat folder on these people - it's so tedious but it's like the last gasp of the dinosaurs. 🦖🦕
It was a story carried by three UK-based media outlets on the following weekend, but they misidentified the threat as an embedded "Kill Switch". See this article by LBC
Ok, that's really old news. It was resolved before it was made public - all the kit (which included Sunamp & other big names) has been patched and sorted. It was never suggested it was 'deliberately installed' either, it was just as likely to be slackness or similar. Who knows?
Be careful of the endless 'Reds under the Beds' narratives that get pushed - there's a huge amount of lobbying behind most of it. And it's under pinned by outlets looking for clicks.
As for Spain.... The first thing they did was investigate for malicious intent - the Spanish MOD was on the government's investigatory panel. Obviously, other European defence 'specialists' were consulted and it was ruled out. It is not an issue.
