Delta M-4, String drops out to zero for 2 hrs ? [delta inverter issues]

[Bozon]

I have a 4.7 kw panel system. Delta M-4 imverter. Systems been live for about 4 months with no issues until now.

The last few days I noticed dips in output on full sun, zero cloud days. One string drops to zero output.

Anyone else have this happen?

I put in service request Monday. Waiting for call back.

 

[SoCalSolarRoof]

That is strange.

Do you have the delta M pro app (android)? If you do, go get the detailed data via Bluetooth.

Admin code is 6532.

 

[Bozon]

That is strange.

Do you have the delta M pro app (android)? If you do, go get the detailed data via Bluetooth.

Admin code is 6532.

I do have it. I confirmed zero output on the PV1 string when it happens.

 

[Bozon]

I do have it. I confirmed zero output on the PV1 string when it happens.

There are some PV2 over current errors the last few days. Their timestamps partially overlap the string drop time.

Error is on PV2...but PV1 is thenone with no output.

 

[SoCalSolarRoof]

There are some PV2 over current errors the last few days. Their timestamps partially overlap the string drop time.

Error is on PV2...but PV1 is thenone with no output.

Interesting. Definitely get them out to check it, over current shouldn't happen, they may not have set the strings up correctly.

 

[jat255]

I have the same setup with the Delta M4 inverter. If you notice, your peak is at 4 kW. I'm almost certain (not an EE, but am an engineer otherwise...) what's happening is that your inverter (and mine, check out my production from today in the image below) is receiving more DC power from the panels than its rated output, which is 4kW. Since my system is 5.44 kW installed in two strings, the only thing it can do is shut off one of the strings to prevent from overloading the inverter (you can see this in my second picture where the western string drops to 0 current over that time frame). This is called "clipping", and is not terribly unusual in solar installations (see How much clipping for east/west solar?). The issue is that instead of the SolarEdge inverter with power optimizers that will reduce output on a per-panel basis, meaning the system can maintain 4 kW when it clips, the only thing the Delta inverter can do is shut off individual strings, which for my system is half it's capacity. Thus, the output at the best solar hours of the day drops to ~2.2kW.

I am extremely disappointed that this is happening, especially since I was quoted a SolarEdge inverter, and didn't realize they installed something else until they were pretty much done the installation (talk about bait and switch...). To me, this is unacceptable, and I've requested that they replace my inverter with a larger size to prevent this from happening (the cost of a 5kW inverter is ~$100 more, so effectively nothing in comparison to the cost of the system). Tesla cheaped out here, and we're paying the price with reduced production. They could have wired it in more strings (the Delta inverter supports up to 4), so it could be more granular when shunting off production, or could have sized the inverter higher to avoid this problem. I have been told I will hear back "shortly" after they've reviewed my request, so I'll try to keep this thread updated with what I hear.

 

[cali8484]

I have the same setup with the Delta M4 inverter. If you notice, your peak is at 4 kW. I'm almost certain (not an EE, but am an engineer otherwise...) what's happening is that your inverter (and mine, check out my production from today in the image below) is receiving more DC power from the panels than its rated output, which is 4kW. Since my system is 5.44 kW installed in two strings, the only thing it can do is shut off one of the strings to prevent from overloading the inverter (you can see this in my second picture where the western string drops to 0 current over that time frame). This is called "clipping", and is not terribly unusual in solar installations (see How much clipping for east/west solar?). The issue is that instead of the SolarEdge inverter with power optimizers that will reduce output on a per-panel basis, meaning the system can maintain 4 kW when it clips, the only thing the Delta inverter can do is shut off individual strings, which for my system is half it's capacity. Thus, the output at the best solar hours of the day drops to ~2.2kW.

I am extremely disappointed that this is happening, especially since I was quoted a SolarEdge inverter, and didn't realize they installed something else until they were pretty much done the installation (talk about bait and switch...). To me, this is unacceptable, and I've requested that they replace my inverter with a larger size to prevent this from happening (the cost of a 5kW inverter is ~$100 more, so effectively nothing in comparison to the cost of the system). Tesla cheaped out here, and we're paying the price with reduced production. They could have wired it in more strings (the Delta inverter supports up to 4), so it could be more granular when shunting off production, or could have sized the inverter higher to avoid this problem. I have been told I will hear back "shortly" after they've reviewed my request, so I'll try to keep this thread updated with what I hear.



View attachment 649830
View attachment 649832

That's some ugly graphs. Tesla must have almost no margin on these projects to cheap out for $100. I hope you can get the larger inverter.

Just to be clear, what you are seeing is not clipping. Clipping is when inverters only produce its max rated output even if the solar panels are generating more. You inverter is not providing rated output and actually in a fault condition.

 

[jat255]

Just to be clear, what you are seeing is not clipping. Clipping is when inverters only produce its max rated output even if the solar panels are generating more. You inverter is not providing rated output and actually in a fault condition.

Yes and no, I think. It's clipping in the sense that the inverter cannot handle over its rated output power (4kW) so it needs to reduce the output, but the only way it can reduce the power is by shutting down one of the strings (since it has no panel-level optimizers). Whether that's by design or faulty, I don't know enough about the Deltas to know.

 

[jboy210]

The more I hear about these Delta inverters we have (4 kW and 8kW), the more question I have about them. I might talk to Tesla and see we can work out a deal to swap them for Tesla inverters.

 

[dareed1]

but the only way it can reduce the power is by shutting down one of the strings (since it has no panel-level optimizers).

This assumption is simply incorrect. Any modern string inverter can reduce the power by restricting the input current on/off times in the inverter circuit. It is the same activity as the inverter does routinely for maximum power point tracking--it changes the duty cycle of the input current.
In the graphs you show, the inverter is not operating properly.

 

[jat255]

They could have wired it in more strings (the Delta inverter supports up to 4), so it could be more granular when shunting off production,

I was incorrect here, it looks like the Delta inverter only supports 2 strings according to the datasheet, so that's all it can do.

 

[jat255]

This assumption is simply incorrect. Any modern string inverter can reduce the power by restricting the input current on/off times in the inverter circuit. It is the same activity as the inverter does routinely for maximum power point tracking--it changes the duty cycle of the input current.
In the graphs you show, the inverter is not operating properly.

Oh, ok! Thanks for the info. I'm out of my depth a bit on how these things work. If it is indeed faulty, I suppose that's "better" for me in that it's more likely I can get it fixed. That's the behavior I've seen in general from this inverter so far. As soon as anything happens to need to reduce power (partial shading, over-power, etc.) it just shuts one string down completely. The installers assured me that the panels are in "groups of three" or something, so partial shading shouldn't knock out the entire string (I have 8 panels on each string), but that has not matched with what I've noticed thus far.

 

[MorrisonHiker]

We had a similar issue with one of our Delta Solivia inverters in 2020 (but not 2018 or 2019). We can't monitor the individual strings (that we know of) but one of our four inverters would completely stop producing. I would notice the issue on the production graph, turn off the inverter, turn it back on and it fixed the issue for the rest of the day. I had to do this several times in the summer of 2020. We reported the issue and Tesla came out and checked the combiner boxes but it was when I was out of town so I don't know if they made any changes. We only had the problem on certain days last summer when it was over 70 degrees. We were having other issues last summer as well which hopefully were rectified when a new transformer was installed.

We haven't had temperatures in the 70s so far this year. The warmer weather next week will be interesting.

 

[Piney999]

Following. I am having the same problem with my Delta M8 inverter. Today is day one, so I am less then happy with things.

 

[MorrisonHiker]

Maybe it's just a coincidence but I've noticed the three users reporting the problem are in Colorado (Denver, Boulder, east of Pikes Peak). I'm in Colorado too. I had a similar issue in the past but it hasn't happened so far this year and was with a different model Delta inverter.

 

[Piney999]

Maybe it's just a coincidence but I've noticed the three users reporting the problem are in Colorado (Denver, Boulder, east of Pikes Peak). I'm in Colorado too. I had a similar issue in the past but it hasn't happened so far this year and was with a different model Delta inverter.

It is a cool, and very intense sun here in Colorado today. Could be a coincidence, or could be a problem with a Colorado shipment or installation.

 

[Ampster]

Cool, intense sun could be as others predicted an over voltage and or current situation coupled with a Delta weakness which goes into drastic clipping mode. I see clipping all the time with my Solaredge but never a string shutdown.

 

[jat255]

Go figure, as soon as I started complaining about it, here's what my power output looks like today:

I'm still not thrilled about how high the DC/AC ratio on my system (1.36) in terms of what that means for clipping, but this is better behavior than what I saw earlier this week where it just shut off one half of the whole system

 

[h2ofun]

Go figure, as soon as I started complaining about it, here's what my power output looks like today:

View attachment 650083

I'm still not thrilled about how high the DC/AC ratio on my system (1.36) in terms of what that means for clipping, but this is better behavior than what I saw earlier this week where it just shut off one half of the whole system

I oversized my inverters to never have these issues. BUT, I was not worried about the extra price

 

[jat255]

I oversized my inverters

Yeah, I wish Tesla would have done that too, but it looks like they just grabbed whatever was in the warehouse on the day of the install. From what I've read, it looks like 1.1-1.2 is sort of industry standard, so 1.36 is kind of on the high side. It's silly to charge per watt of panel, but then peg your production with an undersized inverter (yes, I understand you get more with this setup during non-clipping times than a comparable 4 kW panel system, but why, to save $100 on the inverter and maybe $10 on the smaller breaker?)