Grid Forming Inverters

[magnet creek]

Grid Forming Inverters are being suggested as a solution for the Solar and wind "supply/demand" problem. https://www.sciencenews.org/article/one-device-transform-power-electical-grid-inverter
I will admit that I do not really not what the difference is between "grid following" and "grid forming". I do get it conceptually but I am not sure how this all works in practice. in the article they describe it like this : "A grid-forming inverter can inject voltage into a grid and then adjust its frequency according to however much power is flowing through the system. Other sources of electricity flowing into the grid, the baby ducklings, can then synchronize with that grid-forming inverter, just as they did with the flow of electricity that pulsed from power stations."
I just do not know what is meant by "inject Voltage" and "adjust frequency". In my book, The voltage is 240V and the Frequency is 60HZ.

I know there are a handful of Inverters that can act as Micro Grid Inverters. (SMA makes them I believe). Can somebody explain how grid forming works and if these Inverters are available for the consumer market?

 

[hwy17]

In off grid, grid forming means they can power loads without any grid input.

In utility power, grid forming means what you are reading about. It's mostly a concept for utility scale solar and wind farm inverters.

They don't really overlap, except that some of the same electrical concepts are involved.

Almost all of us in this forum are interested in the off grid grid forming type.

 

[timselectric]

We don't use the terms "grid forming" and "grid following" here.
Because they are confusingly backwards terms.

Grid-tied
connected to the grid source and matches voltage and frequency.
All solar production is pushed (exported) to the grid source. (Main service panel)
Anything that you don't use, is sent to the local grid.
Any power that you need, but aren't producing. Comes from the local grid.

Off grid
An inverter that turns battery power into AC power for local use.

AIO (All In One)
Off grid Inverter, SCC (Solar Charge Controller), AC charger, and a transfer switch. All together in one unit.

Hybrid AIO
Capable of all the above.

 

[cinergi]

Yeah in the DIY world I imagine it's largely irrelevant ... but having watched some power plant startups and reading the article above, (@timselectric correct me if I'm wrong) it's essentially about inverters that can join a grid (follow) versus inverters that can black-start (form) a grid. The inverters we tend to discuss here join the grid (follow) or just provide power to your house (off grid).
While 60 Hz is 60 Hz, you must be phase-synchronized to that to join it... whereas "forming" a grid means YOU set the wave form and others join you. But again, really only relevant in actual grid-scale (including micro-grid) applications.

 

[timselectric]

Yeah in the DIY world I imagine it's largely irrelevant ... but having watched some power plant startups and reading the article above, (@timselectric correct me if I'm wrong) it's essentially about inverters that can join a grid (follow) versus inverters that can black-start (form) a grid. The inverters we tend to discuss here join the grid (follow) or just provide power to your house (off grid).
While 60 Hz is 60 Hz, you must be phase-synchronized to that to join it... whereas "forming" a grid means YOU set the wave form and others join you. But again, really only relevant in actual grid-scale (including micro-grid) applications.

Actually it's the opposite. (Zero common sense to it) Which is why it's confusing.
And why it's best to avoid those terms.

 

[zanydroid]

Yeah in the DIY world I imagine it's largely irrelevant ... but having watched some power plant startups and reading the article above, (@timselectric correct me if I'm wrong) it's essentially about inverters that can join a grid (follow) versus inverters that can black-start (form) a grid. The inverters we tend to discuss here join the grid (follow) or just provide power to your house (off grid).
While 60 Hz is 60 Hz, you must be phase-synchronized to that to join it... whereas "forming" a grid means YOU set the wave form and others join you. But again, really only relevant in actual grid-scale (including micro-grid) applications.

The grid forming literature AFAICT in utility scale power electronics basically comes up once a year on this forum. This thread being that one time. There’s not really a strong incentive or forcing function to make the DIY terminology comply with that

It covers things like black start of utility and operating a grid that has much lower ratio of mechanically coupled spinning generators. I suspect much of the solar battery and wind capacity is NOT proven to be able to form a grid. Nor was it intended to, that is a problem to be prepared for at the appropriate time.

 

[zanydroid]

Another example of a disconnect between DIY and utility scale is what AC coupling means and why one would want it. I believe it’s enjoyed much more in utility scale than in DIY, where in the latter case the advice is to prefer to DC couple things, and in many cases if the DIY system is designed from scratch that is entirely possible.

While there are likely fundamental restrictions to massive DC coupled only utility scale solar plants

 

[hwy17]

If you go with double conversion than you can always be forming your own grid.

 

[Ampster]

I use the term grid forming in the context of being able to AC couple. Also in discussions of that topic it is convenient to call one Grid Forming and the other Grid Dependant (typically a GT inverter). To me that is more than just what an off grid or hybrid can do. Not all hybrids can do AC coupling. An example was the Outback Radian and Skybox which initially could not do AC coupling. Later, firmware upgrades gave them that capability several years ago.

 

[magnet creek]

A lot of talk about "terminology" and how things are called. I am interested in how this actually really works. How do grid-tied inverters actually "supply" excess power to a grid and how does a grid "absorb" this power? Is the Voltage increased slightly creating current in the opposite direction or how does this work? Is this a managed system on the grid side or is all simple physics where electrons flow the other way if you increase the voltage? I have tried to look it up but all the info I could find simply describes the principle of "feeding back to the grid". Any links to detailed reading material would be appreciated.

I think the concept of micro-grids could potentially have huge benefits and could really expedite the growth of solar. If communities are installing more solar and more battery capacity and we find ways to connect and share all of this capacity we are solving the biggest problem in sustainable energy (supply and demand discrepancy).

 

[timselectric]

Is the Voltage increased slightly creating current in the opposite direction

Yes

 

[hwy17]

Trying to put multiple non communicating inverters together in a microgrid is gonna be like trying to stuff multiple cats into a bag.

It's something engineers at ABB will work on, and we will just really try and avoid.

Kauai has the right idea, keep your turbine spinning for the inertia and then let all the inverters follow that.

I would like to add some spinning inertia to my off grid system, but I still am not sure how to or if it's even possible to do it on single phase only.

 

[zanydroid]

Is this a managed system on the grid side or is all simple physics where electrons flow the other way if you increase the voltage?

There is a managed system sort of in the works. Look up CSIP 2030.5 and Smart Inverter from NREL. Plenty of papers to read. That upcoming generation I believe is API based

The most widely deployed control is based on response to electrical characteristics like frequency and voltage. It is somewhat self correcting in response to undervoltage or overvoltage and under/over frequency that results from too much solar getting pushed to the grid, and its interaction with spinning generators.

And then there are the smart meters monitoring the backfeed, capable of operating at pretty high sample rate and granularity

So these electrical characteristics based systems do involve a control loop/defined distributed behavior, however they may not be what you meant by controlled.

 

[Ampster]

How do grid-tied inverters actually "supply" excess power to a grid and how does a grid "absorb" this power?

The grid presents an infinite load to the GT inverter. Within reason the grid can absorb a lot of power. The reality is that most of the energy exported goes to the path of least resistance and feeds the houses on the same transformer first then the other close by houses on nearby transformers.

Is the Voltage increased slightly creating current in the opposite direction or how does this work?

Yes the GT inverter increases voltage slightly to cause current to flow to the grid.

Is this a managed system on the grid side or is all simple physics where electrons flow the other way if you increase the voltage?

Simple physics. The grid has no management except at substations and that is typically a long way away and the exported power gets used by serving the loads closest to the exporting inverter. In some places like Hawaii the substations have become overloaded because of too much solar. My meager understanding is that the control and management systems at most substations were designed for one way flow and those control systems cannot manage reverse flow even though the transformers are mostly biderectional.

 

[magnet creek]

Thanks for the responses, learned a lot.
It does look like the "grid" could use some kind of communication standard. Managing the capacity in a cluster of around 1500 households (which is a good average substation size) with some hybrid of solar and batteries seems somewhat feasible. If data like battery SOC were shared and combined with weather data you could manage these clusters quite effectively, but currently, the grid operators are in the blind, and the amount of grid-connected battery storage is limited.
It should be relatively simple to superimpose a data signal onto the utility feed; not sure how well this would work over a transformer.... and there does not appear to be a standard protocol for this type of data exchange (or is some communication standard in the works?)
While all these may be beyond the scope of many of us on this forum, you can see that more recent product launches (like the EG4 18kPV) are more and more ready for this hybrid setup where grid backfeeding is a "standard" feature. I expect this to be the norm for most future inverters.

 

[zanydroid]

It should be relatively simple to superimpose a data signal onto the utility feed; not sure how well this would work over a transformer.... and there does not appear to be a standard protocol for this type of data exchange (or is some communication standard in the works?)

You can look at what SunSpec and IEEE has for this, 2030.5 is being rolled out. 1741SB is also mandated in most places now. And there's usually an ongoing standardization process for things 5 years out. So from the standardization discussion: you can look at all the public documentation/slides/discussions etc for the last two iterations - 1741SA and 1741SB and 2030.5's various revision are quite well documented in slides starting around 2013 time frame of publication, and I learned (slowly and painfully) primarily off various slide decks. Didn't keep a bibliography to help others learn though.

For the 5-10 years out stuff. It would also be something power engineering research labs would look at, so go to a graduate-level reading class's paper list and have at it. This would cover simulations and potential control schemes. Power engineering is far from my professional area so I can't point you at a bibliography.

Another angle to start research is to look at recent V2X papers and see what their bibliographies / related work sections have on control schemes.

 

[solar8484]

It's the DIY terminology that's often subjectively defined/used sometimes intentionally for marketing purposes. The standards based on terminology is much more objectively clear. Grid-forming inverters, in the research context, are not just inverters that can power loads without the grid. GFI's can also compensate for load disturbances and control DER's via frequency-watt, volt-watt, volt-VAR, artificial inertia, etc. GFI's are also not exclusive to utility scale systems, they are used in micorgrids as small as a single building.

 

[Ampster]

It does look like the "grid" could use some kind of communication standard.

UL1741 is not a communication standard per se but it determines what GT inverters will do if frequency or volts are out of spec. There are some terms I have seen discussed from the Sunspec Alliance so there are some standards but as far as I can tell they are not implemented. My Inverter has some form of Sunspec capability but i know very little about it.

 

[Aloysius]

My understanding is thatgrid forming inverters are the key to macro and micro grid control and are the hot topic around 1741 SA, but they are not new. the sunny island inverter is almost 20 years old and is a grid forming inverter that can form a microgrid that can control the output of standard gridtied inverters that respond to frequency control

 

[ubombi]

As far as I understand,
Grid forming inverters emulate synchronous generator inertia.