How does anti-islanding work?

Status
Not open for further replies.
Hi,

I have a client who wants to know how anti-islanding works, but he wants an extremely technical response from me. Although I'm an electrical engineer, I only know how it works at a general level: if the inverter senses the grid frequency or grid pressure to shift beyond the acceptable parameters, the inverter will shut down. Would someone be able to offer me an in-depth explanation of the sensing mechanism/s generally used, and/or the solid-state physics behind it?

Thank you!

Kind Regards,
Andrew
 

jaggedben

Senior Member
Location
Northern California
Occupation
Solar and Energy Storage Installer
Tell him that inverters use highly complex algorithms which measure voltage and frequency for very subtle abnormalities. Beyond this, the algorithms are proprietary trade secrets so you wouldn't be allowed to tell him if you were someone who knew. You might find some scientific papers on various methods if you put in some time on Google, but they probably wouldn't be exactly the same as whatever product you might be offering him.

One more general thing to point out is that inverters are configured as current sources, so if they cannot match the connected loads the voltage will not stay stable and they'll shut down when they go outside the window.
 

fmtjfw

Senior Member
Last edited:

ggunn

PE (Electrical), NABCEP certified
Location
Austin, TX, USA
Occupation
Consulting Electrical Engineer - Photovoltaic Systems
Well, part of the reason is that when your portion of the grid goes down, everyone else on it still has their loads connected, so the grid looks like a dead short to your inverter. Your 5kW Sunny Boy is not going to be able to light up your city by itself.
 
Well, part of the reason is that when your portion of the grid goes down, everyone else on it still has their loads connected, so the grid looks like a dead short to your inverter. Your 5kW Sunny Boy is not going to be able to light up your city by itself.

Right. My gut instinct is that 99% of the time, anti-islanding is "easy". The situation that I imagine is trickier is when there is another inverter in the system and it just so happens that the loads on the system are close to the MPP wattage of the array at that moment.
 

iwire

Moderator
Staff member
Location
Massachusetts
Well, part of the reason is that when your portion of the grid goes down, everyone else on it still has their loads connected, so the grid looks like a dead short to your inverter. Your 5kW Sunny Boy is not going to be able to light up your city by itself.

Yet if you open the inveters input breaker the inverter still shuts down with no 'short'.
 

iwire

Moderator
Staff member
Location
Massachusetts
My point was if it was the load of the neighborhood that caused the inverter to shut down it would remain running with an open circuit.
 

fmtjfw

Senior Member
My point was if it was the load of the neighborhood that caused the inverter to shut down it would remain running with an open circuit.

No, the anti-islanding would also trip it out. It looks for interactions with the grid not just an increased load. With my system in MA it would dump essentially immediately (within a couple of cycles max as I understand it) when you opened the breaker.
 
The part in italics, particularly Slip-Mode Frequency Shift Algorithm...

Is certainly enough to baffle me! But that's not too hard. :D

Islanding Detection in Distribution Networks with Distributed Energy Resources before Main Circuit Breaker Opening
http://www.hrpub.org/download/20131215/UJEEE1-14901525.pdf
The frequency of the MG (microgrid) has a strong relation with the active power of the MG in an islanded operation mode. Therefore, the control strategy of the DGs in an islanding operation mode is the frequency and voltage control. The islanding detection methods should detect the islanding condition as soon as possible with minimum non-detection zone (NDZ). NDZ is "the difference between active and reactive power on the island when islanding cannot be detected by the corresponding method" [4]. The islanding detection method can be classified into the remote methods and the local methods. The local methods monitor the network parameters such as voltage, current, frequency and active power at the terminal of DGs or at the point of common coupling (PCC). The measured parameters have a considerable variation when the islanding condition is occurred. The local methods classified into the active methods and the passive methods. The implementation of the passive methods is simple but their NDZ are large. The active methods have less NDZ than the passive methods but cause the power quality problems in distribution network because of injecting the continual disturbances to it. Voltage unbalance [5], rate of change of output power [6], rate of change of frequency [7], active frequency drift [8] and Slip-Mode Frequency Shift Algorithm [9] are some of the main local methods. The remote methods monitor the main circuit breaker operation. The main circuit breaker opening causes the islanding condition occurrence. Therefore, these methods detect the islanding conditions by the main circuit breaker opening. The NDZ of the remote methods is minimum but they are expensive methods and their implementation is difficult. Transfer trip scheme [10] and Power line signaling [11] are two example approach of the remote method.
This paper presents a new method for islanding detection in distribution network. This method is based on the zero sequence angle of the current measured at the utility side of the main circuit breaker. Selecting this point as the measuring point causes the fast islanding detection. The measured angle is processed by discrete wavelet transform to extract the characteristics of the measured current, by which the islanding condition can be detected fast and accurately.
----

And you can find these parameters: The 27, 59, 59N, 81U/O and voltage controlled over current 51C
in some inverter installation manuals.
This is what the POCO around here says.

5.7.10.1.1 Inverter-based PV Generator Equal or Above 500kW
The internal relays are considered as primary protection and the Company
requires one additional utility grade relay to be installed as the secondary
protection if the inverter-based PV size is equal or above 500kW. The 27, 59,
59N, 81U/O and voltage controlled over current 51C (Phase and Ground)
elements shall be activated in the redundant relay. The utility grade relay
failure alarm shall be wired to the Company-designated generator interrupting
device. If the step-up transformer configuration is primary wye grounded -
secondary delta, the over voltage element (59N) shall have the capability of
detecting the Line-Ground faults on the delta side and trip the high side (utility
side) interrupting device to separate the Company’s EPS from the Line-
Ground faults on the delta side (customer side) of the step-up transformer.

5.7.10.1.2 Inverter-based PV Generator Below 500kW
If the inverter is IEEE 1547 compliant and UL-1741 listed, the internal relays
are considered as the only required protection and an additional utility grade
relay is not necessary to be installed.
 

ggunn

PE (Electrical), NABCEP certified
Location
Austin, TX, USA
Occupation
Consulting Electrical Engineer - Photovoltaic Systems
The part in italics, particularly Slip-Mode Frequency Shift Algorithm...

Is certainly enough to baffle me! But that's not too hard. :D

Do you think anyone actually reads these huge blocks of text that you cut and paste in here? I know I don't.
 

ggunn

PE (Electrical), NABCEP certified
Location
Austin, TX, USA
Occupation
Consulting Electrical Engineer - Photovoltaic Systems
My point was if it was the load of the neighborhood that caused the inverter to shut down it would remain running with an open circuit.
And mine wasn't that that was the whole picture, just that it would be very difficult to build an inverter that would keep running while connected to the grid in parallel with all the other loads in town during an outage.
 

romex jockey

Senior Member
Location
Vermont
Occupation
electrician
The active methods have less NDZ than the passive methods but cause the power quality problems in distribution network because of injecting the continual disturbances to it. Voltage unbalance [5], rate of change of output power [6], rate of change of frequency [7], active frequency drift [8] and Slip-Mode Frequency Shift Algorithm [9] are some of the main local methods. The remote methods monitor the main circuit breaker operation. The main circuit breaker opening causes the islanding condition occurrence. Therefore, these methods detect the islanding conditions by the main circuit breaker opening. The NDZ of the remote methods is minimum but they are expensive methods and their implementation is difficult. Transfer trip scheme [10] and Power line signaling [11] are two example approach of the remote method.


I've little clue as to what /how/ when these terms apply....:(

But i do know many grid tie folks also have a genny . So my Q would be, does the inverter know the dif between the poco & the genny? Or would it simply continue to 'help the genny out' .....?

~RJ~
 

SolarPro

Senior Member
Location
Austin, TX
Most generators can't maintain the tight power quality requirements that an interactive inverter needs to see in order to export power. In the event that a generator is large enough to maintain a stable power output, an interactive inverter could interconnect with it. Solar microgrids are designed to work this way. But microgrids are far more complex than the parallel power systems that most customers have at home or at their business.

This article details some of the issues involved w/ generators and interactive inverters: Interaction of Grid-Direct Inverters with Backup Generators

(FWIW, outside of places like Vermont, most grid-tied PV systems do not have backup generators.)
 

GoldDigger

Moderator
Staff member
Location
Placerville, CA, USA
Occupation
Retired PV System Designer
I've little clue as to what /how/ when these terms apply....:(

But i do know many grid tie folks also have a genny . So my Q would be, does the inverter know the dif between the poco & the genny? Or would it simply continue to 'help the genny out' .....?

~RJ~
Different anti-islanding (AI) algorithms will be more or less sensitive to the differences between generator and grid. The genny frequency and voltage would have to be more stable than the typical portable generator to qualify for GTI startup. Hybrid inverters with generator support usually loosen the voltage and frequency specs when in that mode.
The biggest differences will come when the GTI tries to backfeed the generator.
At that point if the AI does not promptly shut the GTI down the generator may be damaged.
 

ggunn

PE (Electrical), NABCEP certified
Location
Austin, TX, USA
Occupation
Consulting Electrical Engineer - Photovoltaic Systems
I've little clue as to what /how/ when these terms apply....:(

But i do know many grid tie folks also have a genny . So my Q would be, does the inverter know the dif between the poco & the genny? Or would it simply continue to 'help the genny out' .....?

~RJ~
It's a non issue because grid tied PV is properly interconnected on the other side of the ATF from the generator and shuts down during a grid outage. Even if the PV would run with the generator supplying the AC voltage reference, it's not a good idea to run them both at the same time on the same loads. If the consumption by the loads drops below the output of the PV, the PV will backfeed the generator with a possibly catastrophic result for the generator. The generator is a voltage source and a grid tied PV inverter is a current source; it's not a good idea to run them both together in a closed system.
 
I've little clue as to what /how/ when these terms apply....:(

But i do know many grid tie folks also have a genny . So my Q would be, does the inverter know the dif between the poco & the genny? Or would it simply continue to 'help the genny out' .....?

~RJ~


So the short answer is- the only inverters that know the difference between POCO and gen are hybrid or on/off grid inverters like Outback 8048s or Sunny Island 6048s.

The 8048s have AC inputs for grid AND generator, the 6048 have 1 AC input.

I talked to an Outback guy on the phone and the 8048s only pass the gen straight thru to loads. The manual says "any AC source can be attached to the gen input of the 8048" but that doesn't include a regular AC PV inverter output- the regular GTI inverters need the "infinite" grid for their output- they can't throttle down like a gen can.
There is a newer model of SMA (XXXXTL-US-22) that turns into a 1500w outlet when the grid goes down, but those don't play with gens either. (or Outbacks)

If I've got it right- an Outback 8048, you can attach DC PV panels (or wind power), a generator, batteries and the grid.
A Sunny Island 6048, you can attach AC PV/grid/wind, a gen, batteries and grid.
The ways they function are different too.
If I remember correctly, the 8048 won't charge the batteries off of the gen, while the 6048 will.
This Two Stack single line is a good example of how 8048s work- the AC bypass switch.


http://www.outbackpower.com/wiring-diagrams


Here's one about 6048s.
http://files.sma.de/dl/5610/TechNDIGIN-HUS123112.pdf

I'm figuring it out too- just trying to help ya out.
The stuff below, 81 O/U is [8], while 59 and 27 are [5]- those protections are all built into grid tied inverters (UL1741 cert. inverters anyway)...
So the GTIs monitor these parameters, and they're adjustable, but the maximum current isn't adjustable- if there's nowhere for max current to go, click.

Voltage unbalance [5], rate of change of output power [6], rate of change of frequency [7], active frequency drift [8] and Slip-Mode Frequency Shift Algorithm [9]

ISLAND.jpg


 
Status
Not open for further replies.
Top