Many inverters are designed to take care of this internally. It sort of sounds like you are asking for an inverter with a dual mode of operation, which I don't know of any.

I know the inverters can synchronize themselves to the grid. I want to know if they will protect themselves in the event that they inverter is running as an island while the grid is off, and then the grid comes back on. The grid will not synchronize with anything, so if the inverter doesn't shut itself off, there will be smoke.

This is a question for the inverter designer.
One possible solution or path to a solution;
Provide a very low current signal from the grid to the inverter, possibly impedance limited to milli-Amps.
Let the inverter synchronize with that signal before closing direct on-line.
Alternately, consider an external contactor and a sync-check relay. You will need to be able to set the inverter frequency slightly off of grid frequency so that it may be able to drift into sync.

--------------------
Ohm's law
Not just a good idea;
It's the LAW!

Most inverters are equipped with anti-islanding feature to prevent the case you are describing OP

You need a grid forming inverter.

https://www.energy.gov/eere/solar/articles/powerin...

Use UPS type inverters.
Charge the batteries directly with the renewable power.
Or use some type of DC tie between the inverter and the grid.
The technology is here,but it may not be off-the-shelf for your particular application and/or scale.


--------------------
Ohm's law
Not just a good idea;
It's the LAW!

It appears that this is an off the shelf product. I just haven't found it yet. Why would anyone have battery back up if the batteries weren't active when the grid was down?

I think this is just called a hybrid inverter.

You're describing an islanding microgrid. When grid connected, the distributed energy resources, i.e., PV, wind, battery energy storage system (BESS) will operate in parallel with the grid. When the grid goes down, a point of interconnection disconnecting device should open to disconnect the grid (this is usually operated by a undervoltage/underfrequency relay function), the BESS should switch from grid-following mode to grid-forming mode to power the island - the island may go black briefly unless the BESS is big enough to handle the full load or a fast load shed may have to occur, but open transition is easier than seamless transition - and the PV inverter will automatically wait until it sees a stable source, then synchronize to the grid formed by the BESS. Getting the wind turbine generator to synchronize across its disconnecting device may be more work depending on the type of generator connections - if it feeds through a voltage-source inverter, the inverter can synch automatically. If it is an induction generator, you may just be able to close the disconnect without worrying about synchronizing.

When the grid returns (as detected by relaying or metering), you may simply want to do an easy open transition back to grid through the black state and just close the point of interconnection disconnecting device, or go for the more complicated synchronized seamless transition across it.

Coordinating all of this, not to mention dispatching the paralleled sources and/or loads appropriately, isn't trivial. You may want a microgrid controller designed to handle all of this.

xnuke
"Live and act within the limit of your knowledge and keep expanding it to the limit of your life." Ayn Rand, Atlas Shrugged.
Please see FAQ731-376: Eng-Tips.com Forum Policies for tips on how to make the best use of Eng-Tips.

xnuke,
Thank you. This is helpful. So I'll need a 59 relay to open the grid tie breaker, and of course that breaker has to have a electronic close/trip option. The BESS has to be able to synchronize and to run as an island (maintaining it's own frequency), and then I'll need a relay contact to detect when the grid has become reactivated (after a black out).

Please tell me if this is a correct assessment of how this will operate:

Scenario 1: Grid is active
Each inverter synchronizes itself to the grid when power is available (through solar and wind)
Each inverter shuts itself down when there is not enough power.

Scenario 2: Grid is offline
59 relay detects loss of grid and trips main breaker, creating an island.
Each inverter goes into islanding mode.

Scenario 3: Grid is reactivated after a trip, while system has been running as island.
Relay detects reactivation of grid.
Relay disables all inverters.
Relay closes main CB
Relay reactivates the inverters.

Other questions...
If there are multiple inverters, will one have to be a master for island mode?
If the inverters are online with the grid, and the grid drops out, how can that be detected? The bus will see the voltage created by the inverters?
If the grid drops out, will the inverters see a ton of load and then just trip anyway?

Thanks for your help

Let me understand you want an inverter to power things, and not open a switch or contactor with the grid so that when the grid comes back it will power everything?
First that is impossible, because the inverter would need to be powering the whole (or part of) the grid. My guess the whole grid is several hundred Terawatts.
The reality is you would have an open switch or contactor with the grid. To bring the grid back you would need to cut off the inverter and close in the grid. At that point the inverter may be able to synchronize with the grid.
That is you need an open dead band to bring the grid back.
The other option is have the inverter run in island mode all the time, and have the transition on the DC side, which would not require an open dead band. The problem is when do you do maintenance on the inverter?

No. Maybe I was not clear. The relay will open and close the breaker.

There are three buildings all to be powered by solar and wind whenever possible. There will be two inverters running all the time, one for solar and one for wind. The inverters will each be load connected into a main panel. The main breaker of that panel will be the utility disconnect. Therefore there are three power sources going to the main panel.

Both inverters will be connected to battery banks, and both will be online whenever there is solar or wind available. Whatever power is not used by the local loads will go to the grid via net metering. That's the gist of the system.

I believe hybrid off the shelf is the answer, two inverters in one box. Inverter 1 gets it's sync signal from the grid and operates in net metering mode. When the grid goes away, inverter 1 detects the island situation and shuts down. Inverter 2 now picks up to power critical loads from the battery after transferring them. Look at some of the inverter manufacturer web sites.

stevenal,
Thanks for your reply. The renewables will be sufficient to run all three buildings, not just critical loads. I have been looking into hybrids, and I do agree that this is the hardware needed.

Quote:

The renewables will be sufficient to run all three buildings

And when the renewables are not sufficient to run all three buildings?
Grid failure during a windless night?
Rectify the grid supply and feed DC to the DC bus at a slightly lower voltage than the renewables.

--------------------
Ohm's law
Not just a good idea;
It's the LAW!

I think we are overthinking this. The renewables will act in parallel to the grid. Think of it as a local source with the option of running as an island. The building loads will be served by the grid and by the renewables. If there's enough renewables, then the meter registers 0kW from the grid. If there is no sun and no wind, the building loads are handled by the grid.

Bold text are my changes to your text.

Quote (eeprom)

Scenario 1: Grid is active
Each inverter synchronizes itself to the grid when power is available (through solar and wind).

IEEE 2030.7 calls this the grid-connected steady state. I encourage you to think of solar and wind acting as negative load rather than generation sources since we can't cause their output to go up on command, only down. The battery should be used for renewable firming/peak shaving/energy arbitrage. The utility may not allow export based on your interconnect agreement, so be cautious of that and store excess generation in the battery or curtail the renewables output when the battery is full.

Quote (eeprom)

Each inverter shuts itself down when there is abnormal voltage or frequency.

Correct.

Quote (eeprom)

Scenario 2: Grid goes offline

IEEE 2030.7 calls this a transition from grid-connected to islanded state (unintentional islanding).

Quote (eeprom)

27/59/81O/81U relay detects loss of grid and trips main breaker, creating an island.

Note that this trip cannot lockout the breaker, it has to automatically reset so the breaker can close when the grid comes back, but overcurrent and similar protective trips should lockout the breaker. Also, inverters will need to have their anti-islanding functions shut them down until they see a stable source. Make sure the island will maintain effective grounding as well once disconnected.

Quote (eeprom)

Each inverter goes into islanding mode.

Only one source should be grid-forming, and when you have intermittent sources like solar and wind, that better be the battery power converter. The other inverters stay in grid-following mode. The grid-forming source regulates the voltage and frequency in the island. You may also need your protection trip settings to change once islanded, as the amount of available fault current will probably decrease dramatically. A controller should be managing the charging and discharging of the battery from the other sources so it doesn't run out of energy.

Quote (eeprom)

Scenario 3: Grid is reactivated after a trip, while system has been running as island.

IEEE 2030.7 calls this an open transition from islanded to grid-connected state (reconnection).

Quote (eeprom)

Relay detects reactivation of grid.
Relay disables all inverters.
Relay closes main CB
Relay reactivates the inverters.

I agree with the first action, but it would be better that a controller do the last three of these actions than try to do them with a relay. However, the relay 27/59/81O/81U should all be used in an output to block the close circuit when any one of these is asserted - make sure its VTs are on the line side of the breaker. When all four clear and a synch check 25 function ensures a dead bus transfer will occur once the island sources shut down, set a timer to ensure the grid has been stable for a desired period, then remove the block close and close the breaker. Also, the battery needs to be brought back online in grid-following mode, and protection settings may need to be reset due to the higher fault currents available when connected to the grid.

Quote (eeprom)

Other questions...
If there are multiple inverters, will one have to be a master for island mode?

Yes, the battery. See above transition to island.

Quote (eeprom)

If the inverters are online with the grid, and the grid drops out, how can that be detected? The bus will see the voltage created by the inverters?

See above transition to island. Inverters have anti-islanding detection built in to turn their outputs off to avoid backfeeding the grid. Once the breaker opens, the battery power converter can be brought back online in grid-forming mode.

Quote (eeprom)

If the grid drops out, will the inverters see a ton of load and then just trip anyway?

They'll trip when the grid goes away. Once the breaker is open, you may need to shed non-critical load if the sources in the island aren't adequate to power the whole load. If the entire load exceeds all of the sources, the battery will shut down causing the PV and wind to shut down, and the island will go black.

I highly recommend a microgrid controller to manage all of these things. Also, a competent microgrid design engineer would be able to help size the system components, perform studies, avoid pitfalls, help with understanding codes and regulations (NEC, IEEE 1547 series, IEEE 2030.7, etc.), assist with the utility interconnection agreement and review of utility system and facility impact studies, etc. If you don't have the competency, don't try to do it yourself.

xnuke
"Live and act within the limit of your knowledge and keep expanding it to the limit of your life." Ayn Rand, Atlas Shrugged.
Please see FAQ731-376: Eng-Tips.com Forum Policies for tips on how to make the best use of Eng-Tips.

The idea of rectifying the grid is an interesting idea. The you have a bus with three dc inputs and one AC to the loads. No more synchronization is needed.

You need a sync check before connecting the inverter in parallel with the grid.
If both the grid and the inverter are running at the same frequency but out of phase, the sync check may hang for a long time before drifting into sync.
Anecdote alert: The frequency of a standby generator in droop mode (the most common mode) varies within the droop range with the load.
I have had standby sets with a load such that the set frequency was almost exactly grid frequency but out of phase.
These sets used a very fast open transfer arrangement that needed a sync check.
When the grid returned it may take several minutes for the set to drift into sync with the grid.
Beware of making very fast open transfers without a sync check, particularly if any large motors are running.

--------------------
Ohm's law
Not just a good idea;
It's the LAW!

Searching https://duckduckgo.com/?t=lm&q=Hybrid+Inverter... pulls up some manuals, one of which contained the panel diagram below.

This Hybrid Inverter (probably most) is designed to sit between the grid and the user, all of the power flows through the inverter, unless the inverter is bypassed (bypass mode). The inverters probably can not be paralleled unless specifically designed for that operating mode.

The inverter controller manages the grid connection, and the battery charging.

Quote (waross)

You need a sync check before connecting the inverter in parallel with the grid.
If both the grid and the inverter are running at the same frequency but out of phase, the sync check may hang for a long time before drifting into sync.

Distributed energy inverters and battery power converters provide the sync check internally - they sense the AC voltage, frequency, and phase on both sides of their AC contactor and they modulate their output to match the line side before closing. That's why they won't return to operation in grid-following mode unless they see an AC source on the line side of their AC contactor. There's usually a delay built into their reconnection to ensure the grid is stable - many manufacturers use a default of five minutes, but it can go as high as ten minutes, so there's plenty of time to sync. However, the sync can happen very fast once they decide to do it; it just takes a few cycles since all they have to do is modulate the switching.

The hybrid inverters linked above look like they're for home systems with low power and low voltage requirements. I was happy to learn about them, so thanks to those of you who pointed them out. I even saw some that say they can take inputs from solar, wind, and batteries, so that's nice. I tend to design the controls for higher power and voltage microgrids for commercial and industrial facilities and campuses, so I didn't know these existed.

xnuke
"Live and act within the limit of your knowledge and keep expanding it to the limit of your life." Ayn Rand, Atlas Shrugged.
Please see FAQ731-376: Eng-Tips.com Forum Policies for tips on how to make the best use of Eng-Tips.

This is relatively straightforward in certain configurations now, but there's still a few caveats, and it's subject to how big the system and load is.
As an example, SMA has components that are capable of being configured to run in either grid-connect or islanded with solar and batteries, they even have manuals that detail the configuration, although that's at domestic load size.
All AC coupled, and the battery inverter is capable enough to regulate the solar inverter's output.

Otherwise xnuke's suggestion of looking at a microgrid controller and getting someone who understands these systems is an excellent suggestion. This (perhaps not with wind as gusting causes unexpected behaviour that doesn't happen with solar) has been done often enough that there's generally an OEM who's worked out the details by now. The architecture (e.g. DC coupled solar plus batteries versus DC batteries and AC coupled solar) might not look like exactly what you expect at first glance though.

EDMS Australia

My utility company has already done this twice.
The original, which was just four battery containers that I was involved with had a dual mode inverter. It was connected in a constant current grid following mode, and had to shutdown and start up in grid forming mode to island a small town. My company did a relatively small microgrid which auto-detected loss of utility power, isolated part of the distribution system and started the battery in island mode.

The new one is a small microgrid with four customers, solar and a battery. I am not quite sure how it works as far as operating in grid forming or grid following modes, I’ll have to ask. I wasn’t involved at all.

Theoretically I believe a grid forming inverter should act very similarly as a synchronous machine. To protect the inverters, like a synchronous machine, you will want to install sync check on the intertie breaker relays to prevent inadvertent out of phase closure with possible damage. If you don’t want to drop and pick you’ll need to install an auto synchronizing device, which can be packaged in a number of different products including common microgrid controller platforms like the SEL RTAC.

Let me know if you need more information.