DC fuse caught fire! Help ?

[KaFun]

I bought specifically this inverter because it can run batteryless and also blend "just the missing required power"

I read through the manual - it seems to be the same as mine. My unit also can run without a battery, but I never tried it. I've found the attached description of the mode you're probably referring to (page 32/60). It seems indeed as if the internal inverter synchronizes its output to the grid. This is actually a great feature and I was not aware of it.
Looking forward to your ongoing tests in this direction. I'd be interested how it would behave if the grid fails and comes back after some time. I'm curious how they would manage to synchronize both inverter and grid with regards to each other. I know from the old APC UPS this to be a pretty complicated thing as it requires some control loops on the inverter side and a relay that adds grid power the second both voltages are in phase.
But when I think about it, this might not happen: If you run your system with loads higher than PV can supply, it would run on both. If you switch off your grid, the whole system would switch off (because PV can carry the load). You then have two options:

1. Reduce load to a level that PV supports --> no problem when grid comes back as it is not needed in that second and the internal inverter can synchronize before a relay switches the grid to the output
2. Wait until the grid is back --> also no problem as the internal inverter might try to carry the load but will always fail.

I'd be interested if this assumption is correct.

 

[Hedges]

I read through the manual - it seems to be the same as mine. My unit also can run without a battery, but I never tried it. I've found the attached description of the mode you're probably referring to (page 32/60). It seems indeed as if the internal inverter synchronizes its output to the grid. This is actually a great feature and I was not aware of it.

Seems like a good feature.
But if these are not UL-1741 grid-interactive inverters, I do wonder how much they might back feed into the grid.
They likely use current transformers or other sensors to ensure they only feed household loads and don't backfeed the grid.
But sensitivity and accuracy of those may not be as good as GFCI, which trip at a threshold below what is dangerous to life.
So I wonder if it is possible for them to backfeed the grid during a grid failure, and be a hazard to utility workers.

I use an inverter (Sunny Island) which is UL-1741, and can supplement power from the grid. It can also backfeed the the grid if so configured. It performs active anti-islanding when grid is connected.

I'd like to see a certification (UL Listing in the U.S.) for inverters that do blend their power with the grid.

 

[EliteSolar]

Seems like a good feature.
But if these are not UL-1741 grid-interactive inverters, I do wonder how much they might back feed into the grid.
They likely use current transformers or other sensors to ensure they only feed household loads and don't backfeed the grid.
But sensitivity and accuracy of those may not be as good as GFCI, which trip at a threshold below what is dangerous to life.
So I wonder if it is possible for them to backfeed the grid during a grid failure, and be a hazard to utility workers.

I use an inverter (Sunny Island) which is UL-1741, and can supplement power from the grid. It can also backfeed the the grid if so configured. It performs active anti-islanding when grid is connected.

I'd like to see a certification (UL Listing in the U.S.) for inverters that do blend their power with the grid.

That's the only post I found about how these inverters work ( by member marsal505 but it doesn't say if it backfeeds really , just how it works ) :

I was able to study the circuits inside one of these units and this is what I found:

1. when the AC bypass mode is enabled, the AC input and AC outputs lines are bridged together via relay switches

2. the inverter feeds the energy created from the solar panels on to the bridged AC input/output lines by raising the AC voltage via internal semiconductor power switches (IGBTs), in the same way that a normal grid-tied feed-in tariff inverter does. The only difference is that this inverter can never feed energy back into the grid. This must be limited in the software, even though in hardware it is possible to do.

3. this means that when the AC bypass mode is enabled you can go beyond 5KW load - I have successfully gone to over 8KW for several minutes without shutting down - just got the over load beeping.. The maximum power limit would be set by the inverter's internal 40A circuit beaker (about 40A x 250V = 10KW) or what ever circuit breaker you may use on the external AC input line.

4. you are only limited to 5KW maximum load output when you are operating the inverter with a battery, without any AC input connection (ie. when no AC bypass is enabled). The 5KW limit applies for solar + battery together and is the limit of the inverter circuit section, consisting of the IGBT power switches

What would be the cheapest and most reliable way to see if it ever backfeeds to the grid ? I cannot backfeed to the grid.

Btw my inverter doesnt have an option to backfeed the grid.

 

[Hedges]

By setting whatever configuration it has for zero export?
(But without UL Listing, I can't be confident it safely does what it is supposed to.)

We think utility meters in the U.S. connect to only L1 and L2, not N.
So they combine magnetic field of L1 & L2 current, and use voltage between L1 and L2, to measure power.
That means, if V(L1) and V(L2) are different, measured power is not completely accurate.

My Sunny Island has internal current sensor, measures what current is backfed to main panel (and on to loads or back to grid.)
Many inverters use current transformers to measure current from/to grid. People report some amount of backfeed on utility bill, even though they configured for a small (e.g. 30W) import. Maybe a greater minimum import would eliminate backfeed.

My concern regarding safety is that > 30 mA at 120V is considered hazardous, and zero-backfeed configurations might let such a small amount through. And not shut off in human-safety milliseconds rather the multiple seconds of grid-interactive. Might even use utility transformer for balancing, export 1A on L1 and import 1A on L2, while driving high voltage to other side of transformer.

 

[Mattb4]

...

My Sunny Island has internal current sensor, measures what current is backfed to main panel (and on to loads or back to grid.)
Many inverters use current transformers to measure current from/to grid. People report some amount of backfeed on utility bill, even though they configured for a small (e.g. 30W) import. Maybe a greater minimum import would eliminate backfeed.

...

There is a problem about being in parallel and thinking you can eliminate back feeding. Every load on the grid is being carried in proportion by the power supplies that are in parallel. As overall loading changes the power supply has to increase or decrease its output to keep its proportion. It is not just the house loads that affect your inverters output. Standard GTI do not pretend to limit supply to the house and not the entire grid.

 

[EliteSolar]

There is not setting since the model cannot be set to backfeed.

My concern for backfeeding is its illegal to backfeed and I will have 'em for inspection I can even get fined. I have a Smart (digital) Meter.

What would be the cheapest and most reliable way to see if it ever backfeeds to the grid ?

 

[Mattb4]

There is not setting since the model cannot be set to backfeed.

My concern for backfeeding is its illegal to backfeed and I will have 'em for inspection I can even get fined. I have a Smart (digital) Meter.

What would be the cheapest and most reliable way to see if it ever backfeeds to the grid ?

Does you unit work when you remove grid power (open the supply breaker) and have no battery?

 

[SpongeboB Sinewave]

One of my plug-in power meters can show negative watts. IF the system is on the edge of charging and selling to grid, then almost any "decent" power meter should be able to work with this. Wouldn't want to do this with the grid putting out full current, but when near to balanced and up to 20 amps or so, it should work. This one is a HOPI and the reason I bought it was because I saw a tear down of it on the bigclivedotcom YouTube channel. No data logging for most of these but you could watch it or video the screen when it's time to look at it.

boB

 

[EliteSolar]

Does you unit work when you remove grid power (open the supply breaker) and have no battery?

Yes, it can run just on solar panels and output AC directly

I have a watts meter but I dont know if it can show negative watts. Would this be a reliable way to test that the unit never backfeeds even for a ms ?

EDIT: I live in EU , my grid is N L1 , 240V

 

[Mattb4]

Yes, it can run just on solar panels and output AC directly

Have you tested it?

 

[EliteSolar]

Yes, 1st time I disconnected the battery while it was running , the next time it switched on by its own when there was enough light. ( NO AC IN )

 

[Mattb4]

Yes, 1st time I disconnected the battery while it was running , the next time it switched on by its own when there was enough light. ( NO AC IN )

View attachment 115105View attachment 115106

If it can do that that it likely does not grid tie. Any grid tie setup has to disconnect once the AC supply side goes away. No AC signal no back feed.

BTW, it is rare that one of these can function without either AC or Battery. Would be intersting to see how it was wired internally.

 

[EliteSolar]

yes very interesting model , thats why I bought it.

I understand you are quite sure it can never backfeed . But can you please also suggest the best most reliable ( cheap too if possible ) to test if the unit ever backfeeds ? I dont want to be illegal and backfeed ever

Im still in fear , it can backfeed through AC IN .

I will use this model with AC IN to blend the power with the inverters output , I don't know if you have read the thread

I might be able to open it and take pictures if you want but it would be nice if I could avoid it that

 

[Mattb4]

Never said I was sure it could not back feed. Testing for it would involve doing things I simply would not suggest for somone that was not trained in working with electricity.

 

[EliteSolar]

Cool thanks.

Does anyone else know if these type of inverters backfeed to the grid, through AC IN ? There is no setting for exporting , but it can blend/mix AC power with the inverters AC ouput to meed load demands.

Could someone suggest a reliable and cheap method to test for backfeeding ?

I think we can close this thread since the discussion is no longer relative to the original thread title.

I will make a new thread about this but if you still want to contribute here its fine!

Thank you all for your answers and tips

 

[SpongeboB Sinewave]

To know which way the power is flowing, you need to know if the AC current is in-phase with the AC voltage or out of phase with the AC voltage.
That is, 0 degrees in phase or 180 degrees out of phase. You WILL get some phase relations that are NOT these 2 values when below 1.0 power factor is being drawn from or sold to the grid.

That is why a power meter that shows real watts is an easy way. When power factor from/to the grid is 1.0, or close to it, you can look at AC current only when you know if your system is taking power from the grid.. That is, charging or power a load on the home side.

When that power from grid is known to be coming DOWN, then the grid current should come down. Then, when it goes to zero amps, then no selling OR buying to/from the grid. At that point, if you see the AC current going UP again but you know you are not charging or drawing power from the grid to your loads, the it might be selling.

It's a question of balance. And if you have battery current polarity visibility available, you can also get a sense of inverting and charging. If the inverter is inverting and your loads aren't loading more, and the AC current starts rising... Then you might very well be selling from battery or PV.
The PV current of course should be like battery current for inverting but has only one polarity... Charging the battery of selling.

If you do not have a battery in the system, then just need to look at PV current and then you also know whether or not you are inverting or selling.

Just part of some more tricks you can use if the information is available to view but you kind of have to be there and watching things go up and down and be aware of house loads so you can tell if that is where power is going.

boB

 

[EliteSolar]

There is no 2 way watt meter ? Identify the current flow and show me " +500Whr -22Whr " ?

 

[KaFun]

Does anyone else know if these type of inverters backfeed to the grid, through AC IN ?

I don't know, but I have an assumption about this that it definitely won't back feed and here it goes:
Let's assume the AC IN has just two ways to go into the system: Through the bypass relay and through the battery charger (which is actually just an AC/DC converter to a DC rail plus the battery charger which feeds from this rail).

Now there are three cases for running without a battery:

1. load is below what PV can provide
2. load is above that but below AIO power rating and
3. load is above AIO inverter power rating (in your case the 3.6kW).

This would have the AIO do the following:

1. If load is below what PV can provide, the inverter runs on PV only - that's the easy part.
2. If load is above PV but below AIO inverter power rating, the AC IN provides the power difference needed via the AC/DC converter in addition to the PV power - not the most efficient way, but it would work.
3. If load is above the AIO inverter power rating (or PV is off), the system switches to Bypass mode.

Now for the question how to test this:
IMHO the only way to test my hypothesis is to run your system with a load above PV power but within AIO inverter power rating (case 2). Using a power meter in front of the AIO should give you a slightly higher power draw as if you'd connect the load directly to grid AC (as the AC/DC converter loss will be missing).
I guess, a hair dryer running as load on a partly cloudy day will do ...

Now this is a wild speculation, but could answer your question. If there is no difference in power draw, the AC IN would directly feed the AC OUT instead of running AC->DC->AC. And then there might be a chance, the system potentially could back feed ...

 

[SpongeboB Sinewave]

There is no 2 way watt meter ? Identify the current flow and show me " +500Whr -22Whr " ?

Sure there are. But I was thinking more about instantaneous power flow or very small amounts of energy. Very small amount of net kW-Hours more like kW-seconds.

+/- 22 Whour would be quite a bit in the circumstances I am thinking about. When the system is CLOSE to balanced, or +/- 0 amps, it might have a hard time keeping it on the negative (charging) side of that polarity and wander over the line at very low currents.

Probably why it is a good idea to have some kind of loads on to help keep that wandering at a minimum.

In Hawaii, Rule 14H allows selling back to the grid like, 20 or more times per day but I don't know what they mean by that in kW-Hours.

I am also not sure what a lot of utility watt-meters do when you sell back, say, just a few cycles of power to the grid when things wander.
But, just low power factor appliances sell back parts of cycles of AC. Just no net energy above (or less than) zero.
Do some meters use their data over power line or wireless signaling to inform the utility of some certain level of sell-back ?

If the system does not have a CT on the main utility AC distribution panel, then it can't tell if it is selling back or not. It can only tell if it is selling back through its AC input terminals.

The +/- depends on your definition of buying or selling. Usually it is negative when drawing from the battery and positive when charging.
SMA Sunny Islands are opposite but I'm not sure why.

boB

 

[MaximusAnonus]

Hager is (pardon my French) shit. I bet it failed purely because if that. Next time buy some quality gear (I know I know some of these things have loooooooooooooooooooong lead times these days but still...).