25

We had some solar panels & an inverter installed recently. Every time the inverter is turned on (after maintenance or with the first light of the day) it has to synchronize its generation phase with the grid, and it is easy to understand.

But why does it take around 5 minutes? When I asked the guy installing it, he said "it is some sort of standard, every inverter takes around 5 minutes to sync" and had no idea about the reason.

The inverter has 60 changes per second to sync with the grid. Sometimes it might be needed to observe the grid but still, 5 minutes is too much time and missed opportunities to synchronize:

60 Hz x 60 seconds x 5 minutes = 18,000 cycles

Am I missing something?

My inverter is Delta SOLIVIA Solar Inverter 3.8 TL.

Peter Mortensen
  • 1,676
  • 3
  • 17
  • 23
dvdmn
  • 615
  • 2
  • 12
  • 20
  • It's not clear how you expect us to answer this without a make, model and link to specifications for your inverter. – Transistor Jun 18 '18 at 18:26
  • 4
    it seems a common behavior for all inverters. – dvdmn Jun 18 '18 at 18:28
  • 3
    Just guessing, but might be some sort of timeout to prevent a faulty inverter from continually connecting and disconnecting itself from the grid in a way that causes a rapid alternation of power flow on the grid connection. – Jon Jun 18 '18 at 18:37
  • 9
    It is possible "the guy installing it" is less than fully informed about the performance of adaptive-filter PLL and Fourier analysis synchronization techniques? Seems the actual lock time should be less than 1 second. – Spehro Pefhany Jun 18 '18 at 18:39
  • 2
    @Spehro True, but you're assuming the grid (or your local connection to it) is running properly in the first place and doesn't have anything loading it which will mess around with phase and frequency. What's a safe assumption for your domestic supply is less so in more remote rural areas, especially in poorer countries where the infrastructure may not be so good. – Graham Jun 19 '18 at 01:52
  • while we were having the discussion this topic came up and the technician said "this inverter cannot work without grid power. It needs an AC signal to sync itself." We also have a powerwall so inverter has 'an AC signal' to sync itself. Thus I can shut down the grid and I can go off-grid entirely. – dvdmn Jun 19 '18 at 02:42

3 Answers3

38

It's nothing to do with synchronization. It has to do with ensuring safety of utility workers. The inverter should be quick to disconnect in the case of a grid failure (seconds) wait a period of time (in this case 5 minutes) after the grid is restored before beginning to supply power out to the grid.

See, for example this exchange (the "standard" in question is UL 1741/IEEE 1547).

Question:

I understand the requirement to cease energizing within 2 seconds. How long does the inverter have to be de-energized before coming up to provide power to a local load w/ the utility breaker open?

Answer:

The standard does not directly address it, but from a lab perspective, common practice is to have the inverter wait a minimum of 5 minutes after the Area EPS steady-state voltage and frequency have been restored.

Further comments indicate that some European countries require 3 minutes and Australia requires 1 minute.

Spehro Pefhany
  • 376,485
  • 21
  • 320
  • 842
24

The inverter has a software delay. This is intentional.
It waits 5 minutes to make sure it is connected to a stable grid.

An inverter could theoretically connect and go full power in seconds. But it doesn't. For example, if after a power failure, all inverters immediately went online a started outputting full power, the network would be overwhelmed and will fail again due to overspeed.

Instead, it waits for a stable mains connection. Since it can be less stable after a large scale failure, when there is still a lot of switching going on.
And then slowly ramps up the power in a controlled slope.

Some regions have specific regulations for inverters, see your local grid code.

Jeroen3
  • 21,976
  • 36
  • 73
  • 1
    This doesn't explain the delay though. If they wait for a _stable_ connection, it should take much less than 5 minutes for OPs situation because there has not been an outage. And if _every_ device waits 5 minutes after an outage, you still have the same problem - everything comes on at once, just with a 5 minute delay. – pipe Jun 19 '18 at 08:42
  • 1
    @pipe From the inverters point, there was an outage. The inverter cannot know anything about the status of the grid while it was turned off. The inverter is turned on, checks the grid connection for the first time and then waits five minutes. – Josef Jun 19 '18 at 08:54
  • 1
    @pipe There is no real definition of "stable". But if the voltage dips or swells beyond nominal, the timer is reset again. Power restoration isn't as immediate as you might think. Yet it's still a problem if inverters went all-in immediately when you could be connected to a reduced grid. The same applies for emergency generators. They do not transfer load back to mains until it's back up for a while. And even after they keep running for cooling and say 15 minutes extra ready to take over if the outage continues. – Jeroen3 Jun 19 '18 at 09:26
  • 8
    @pipe It's a sensible strategy for small microfeeders. Other, larger, systems obviously have more sophisticated control systems. If you've got 5kW of solar on your roof you're not going to install an industrial grade SCADA system and have an engineer bunk in your guest room full time so waiting five minutes is a reasonable alternative to that. – J... Jun 19 '18 at 15:31
1

As per my observations and experience, 30 to 60 seconds time is sufficient to synchronization of solar grid inverter to connect with grid and export power to grid. The time 30 secs to 60 secs are required for monitoring grid voltage, frequency and phase and estimate angle i.e to satisfy phase lock loop function to sync with grid.

bhima prasad
  • 11
  • 1