Inverter swap speed

[Pocster]

I’ve no issue with my inverter .

But as most people new to pv / battery etc. find ; it can take time to swap .

e.g a sudden draw can cause it to go to the grid before using pv or battery .

Just wondering if anyone knows why this is so ? - for me it can be around 5 seconds or so . Is this just a limitation of current inverter technology ? . Bit like the ac to dc conversion costs 10%. When will these inefficiencies be fixed by more modern technologies?? 😁

Edited April 14, 2023 by pocster

[Temp]

They could probably eliminate the delay by running the inverter constantly but that might waste more energy. 

 

I suppose some types of load might also cause the inverter to switch on and off rapidly if the response time was too quick. I suppose that might put a lot of noise on the supply? 

[Pocster]

16 minutes ago, Temp said:

They could probably eliminate the delay by running the inverter constantly but that might waste more energy. 

 

I suppose some types of load might also cause the inverter to switch on and off rapidly if the response time was too quick. I suppose that might put a lot of noise on the supply? 

What about the ac to dc inefficiently ? - 10% is quite substantial 

[Temp]

6 minutes ago, pocster said:

What about the ac to dc inefficiently ? - 10% is quite substantial 

 

It is when on load but I was thinking more about the losses running the inverter in a standby mode waiting for a load. 

[Nick Thomas]

My inverter also takes ~five seconds. It has a register for "Fast zero export", but doesn't permit it to be set.

 

Regulatory stuff, maybe?

 

 

[Pocster]

12 hours ago, Nick Thomas said:

My inverter also takes ~five seconds. It has a register for "Fast zero export", but doesn't permit it to be set.

 

Regulatory stuff, maybe?

 

 

Yeah could be regulatory stuff I suppose.

[SteamyTea]

I have a vague recollection that non 13A loads i.e. anything hard wired, has to disconnect i.e trip or blow a fuse, within 5 seconds.

This swapping over from grid to local storage may be a legacy of those rules.

We tend to think that microprocessors, running in the gigahertz speed range is all that is needed. Probably find that power electronics cannot run at those frequencies reliably and sensibly.

I may just be talking bollocks, but if it was simple, it would have been done.

[mattman]

Are you referring to "trickle" grid import, when there is plenty of battery energy stored up and/or plenty of PV/sun - yet there is a small delay before the inverter uses either the battery energy and/or PV and instead is using the grid for the first few seconds?  I think this is called "ramp up" times when a load is first noticed by the inverter it takes a very small amount of time to start using either battery or PV power - and during this time the grid is used.  There is a guy on youtube who talks briefly about this, EV Puzzle I think was his channel.  It could be to do with load balancing with grid-tied inverters.  It didn't appear to happen as much when he tested a Huawei inverter compared to another inverter (I can't remember the other brand) but the Huawei had much less of a ramp up time (less trickle grid import) compared to the other one.  I believe Solax inverters also have a way of minimising ramp up times too.  Another alternative is that all house loads run through the (EPS backup mode) batteries first and only when the batteries are nearly depleted or the load is too much for the inverter/batteries' maximum discharge rate only then does the grid temporarily cut in and meet that load demand - he called it a "grid parallel" set-up and Victron is good with this kind of set up although I see no reason why any hybrid inverter cannot also be set up in this way - as long as there is enough battery storage and the fuse on the EPS backup load is big enough.  I am no expert on this but this is all from memory.