What I'm learning about DIY Enphase and Hoymiles Microinverters

[svetz]

... I couldn't care less that the Enphase inverters are slackers....

Having the battery energy (kWh) be balanced would allow for maximum power (kW) at night should you need it. Since they're simple tests, I'd also use the clamp meter to validate some of the assumptions to ensure no bits of Scenario #2 or #3 are occurring.

 

[ncsolarelectric]

Scenario #3 is falsified by the SENSE monitor. Enphase monitor is only 1 decimal place for 100W shown as 0.1kW. SENSE is showing 110.9W. So that's as accurate as it can be.

Scenario #2 however is a good idea for a test and see if when the Enphase inverters show nothing or 6W, verify current is not backfeeding to them.

I realized that if the microgrid were formed by an AIMS Low-frequency transformer inverter, rather than the Enphase IQ8s, all of these issues of potential inverter damage pretty much go away. The AIMS can handle a 3X surge and the way they are designed, any output surge will get pushed back to the battery without damaging the inverter MOSFETS. With the zero-export control on the HM system, it would be a safe and effective means of creating a large distributed microgrid.

 

[ncsolarelectric]

I just ran the HM and the IQB3T together as a microgrid. It worked great! I reduced the load to practically zero. The only load was the Envoy and the refrigerator intermittently, as well as the IQB3T charging itself.

The IQB3T supplied a little power at first with my air conditioner on, and when I turned it off the battery charged to full. Then it did its little on/off cycle for an hour as usual when it charges. During the time in between, I saw the HM inverter throttle all the way back to zero. It didn't trip off-line, there was no alarm fault at that time. When it began to droop, the frequency went up to 60.31 Hz. When it was at 0, the voltage and frequency were normal, 240V and 60Hz.

However, there were two instances where it had a grid over-frequency alarm during the test. The first one when I turned off the grid, and the second one when I turned the grid back on. Again, I think this is the Enphase system phase-locking.

I think this concludes my live field test of the HM-350NT. I'm going to remove it and put the IQ8+ back, then start building my addon system.

 

[ncsolarelectric]

Here is some information from HM tech support regarding grid profiles and operating in parallel with the Enphase system I've been testing. I let them know the results and that there appears to be no issues. In response, he told me how to edit the grid profile, so I can match it to Enphase!

"Hello Todd,

Thank you for all this valuable feedback.

In fact, since we have not performed official specific tests for the specific third-party inverters, we cannot promise full compatibility for related designs so we recommended for our on-grid solutions which will be more mature in this situation since we would also need to cooperate with various settings for other inverters.

Anyway since you have tried this solution with acceptable results, we will provide the information you need for you. Actually, for the drop slope under frequency settings, this value can be adjusted in the grid profile by editing and generating a new one (check the screenshot below, it can be edited between 16.7 to 100), attached you may also refer to this technical note for related applications.


As for the Wattnode meter, since we have no idea whether SolarEdge has some internal settings on their Wattnode meter, we would still recommend using it from CCS to ensure they have factory settings to be compatible with our system.

I hope this information helps with your installation.

PS: Since current applications could be an abnormal operation for our microinverters, in the case that any issues are caused by this application, I'm afraid we may not provide corresponding support or warranty claims related to the compatibility inquiries with these inverters, thank you for your understanding!"

I'm disappointed about the warranty statement, but I can understand that. You know, it's the same with solar panels. If you read the fine print on the warranty. When you mix solar panels from different manufacturers, it voids the warranty, yet people do it all the time.

 

[svetz]

...When you mix solar panels from different manufacturers, it voids the warranty, yet people do it all the time....

Just an FYI that with microinverters the panels don't interact, each is an isolated island. So, can be no "mixing" to void the warranty (not that the warranty claims department minion will understand that or validate actually validate the warranty).

 

[ncsolarelectric]

Here is my next experiment. An ESS using these microinverters.

 

[svetz]

Here is my next experiment. An ESS using these microinverters.

An IQ3 battery has about the same energy (e.g., 3.36 kWh Vs. (3.84 kWh @ 90% DoD=) 3.4 kWh) and is about $3,000... How's the pricing compare?

 

[svetz]

Here is my next experiment. An ESS using these microinverters.

Thoughts:

• I see the battery charger in the schematic, but not on the BoM.
• Assuming the battery has a BMS built-in as the BoM doesn't have one
• You'll probably want a DC cutoff (suggest a breaker) and a battery shunt
• You might want a DC switch with relay so the battery only discharges when you want.
• You need a way to break the charger/inverter loop (otherwise they'll be on at the same time), ideally you want the battery to charge with excess solar
• Typo in the system size:

Seems to me there's a fairly new NEC code spec for ESS, mainly I know about it from threads talking about it makes it much harder to DIY. Might be worth reviewing for compliance if North Carolina has already adopted it.

 

[ncsolarelectric]

An IQ3 battery has about the same energy (e.g., 3.36 kWh Vs. (3.84 kWh @ 90% DoD=) 3.4 kWh) and is about $3,000... How's the pricing compare?

Inverter with fuses, wires, cables, and MC4 connectors is just under $400
The best price I've seen for the CSS Meter + (2) CTs is $559, but if it's the same as SolarEdge's meter, then it's around $200.
However, if I use a charge controller and count the inverters as part of the PV system which can export to the grid, then the zero-export meter is not required at all. It's only required to peak-shave usage from the grid without exporting to the grid from the battery.
An AIMS 40A MPPT charge controller is $342 and it works for a 36Vdc battery, but an AC charger is probably under $200.
That comes to somewhere between $1700 and $2400. At the low end, it's around $445/kWh. If you can find cheap batteries, the cost is up to you.
All I know is that 99% of the time, I use my IQB3T to offset grid usage by peak-shaving what I would take from the grid. I have not needed it as a backup for a grid outage since I installed it. That doesn't concern me, but saving $ on usage does.

 

[ncsolarelectric]

Thoughts:

• I see the battery charger in the schematic, but not on the BoM.
• Assuming the battery has a BMS built-in as the BoM doesn't have one
• You'll probably want a DC cutoff (suggest a breaker) and a battery shunt
• You might want a DC switch with relay so the battery only discharges when you want.
• You need a way to break the charger/inverter loop (otherwise they'll be on at the same time), ideally you want the battery to charge with excess solar
• Typo in the system size: View attachment 105289

Seems to me there's a fairly new NEC code spec for ESS, mainly I know about it from threads talking about it makes it much harder to DIY. Might be worth reviewing for compliance if North Carolina has already adopted it.

Thank you. I always need a second pair of eyes on a new draft revision.

My little drawing is of the EG4 or SOK type 5.12kWh, 48Vdc battery bank because they are the lowest cost for a 48V system. They have a BMS and communication so a shunt isn't necessary, but a nice add-on. I like to look at the basics first.

NEC 2020 supposedly takes effect in NC in a couple of months. It requires any energy storage system over 1kWh be UL Listed, and installed by qualified personnel. I am qualified, not licensed. I've had decades of training. The parts I've specified are all UL Listed, so the system should be compliant. I'm sure the UL testing of the inverter did not use solar panels to do the long-term thermal testing.

 

[ncsolarelectric]

Creating my own grid profile took like 30 seconds. I'm going to compare how the inverters are sharing the load tomorrow, before and after I change the %Pn/Hz droop slope. It is currently set to 50%, I can increase it to 100%, but I've set it to 90% to be "safe". LOL!

The previous test showed that the IQ8+s drooped power much faster and shut down while the HM inverter powered the entire load. I think if i can match the Enphase droop slope, that I can get them to share the load as @fromport suggested.

Here's a dangerous toy. The anti-islanding feature can be disabled. I wonder if it can operate off-grid if that is disabled? Wouldn't that be nice! Experiments abound. Next!

 

[Ampster]

The anti-islanding feature can be disabled. I wonder if it can operate off-grid if that is disabled? Wouldn't that be nice!

Off grid? I thought all your experiments were with the grid disconnected? If the grid was connected, there is nothing to prove except that they work as they are supposed to as GT inverters. What was the microgrid you spoke of?

 

[ncsolarelectric]

Off grid? I thought all your experiments were with the grid disconnected? If the grid was connected, there is nothing to prove except that they work as they are supposed to as GT inverters. What was the microgrid you spoke of?

off-grid, as in, without the Enphase IQ in parallel. I have been testing off-grid but the Enphase IQB3T is creating the microgrid. I'm wondering if disabling anti-islanding will allow the HM to form their own microgrid, like an Easy-IQ8.

 

[Ampster]

This thread started out looking like an interesting insight into Enphase micros. It is evolving into a thread where it looks like you are promoting a product you are selling.

 

[fromport]

off-grid, as in, without the Enphase IQ in parallel. I have been testing off-grid but the Enphase IQB3T is creating the microgrid. I'm wondering if disabling anti-islanding will allow the HM to form their own microgrid, like an Easy-IQ8.

You really think that disabling anti-island functionality is what you need to create a grid ?
I am seriously disappointed.

A grid tie system is 'simply' a current pusher against the existing power cycles.
It can *never* generate a sine wave on its own.

 

[Ampster]

It can *never* generate a sine wave on its own.

Not to mention that it goes through serious tests to confirm that a grid actually exists.

 

[ncsolarelectric]

You really think that disabling anti-island functionality is what you need to create a grid ?
I am seriously disappointed.

A grid tie system is 'simply' a current pusher against the existing power cycles.
It can *never* generate a sine wave on its own.

I know all about it dude. I was designing sine wave inverters in 1988, and I designed a bidirectional inverter that could switch from voltage mode to current mode automatically, or with the press of a button in the 1990s. Neither of us has any idea what the HM inverter can and can't do. I question why they would provide a software switch to turn off anti-islanding, if it wasn't capable of forming an island microgrid in the first place.

 

[ncsolarelectric]

This thread started out looking like an interesting insight into Enphase micros. It is evolving into a thread where it looks like you are promoting a product you are selling.

Hey, I like to tinker with stuff. Right now, I'm tinkering with an HM inverter. I think some folks around here are interested in seeing how these experiments turn out. It's a live review of a new product, with some abnormal tests to learn what happens. I'm an old-school engineer learning about the latest stuff and sharing what I learn. If that doesn't interest you, you don't need to follow it, but you also don't need to be rude. Thanks.

 

[Ampster]

If that doesn't interest you, you don't need to follow it, but you also don't need to be rude.

I did not intend to be rude, I was just trying to give you a hint that you were getting close to the forum rules about vendors, before you got warned by a moderator. Have you read those rules? You have mentioned several times that you are a distributor of that product. That is my definition of a vendor.

Also I am not interested in another thread about someone trying to trick a GT inverter to run off grid so I have unwatched this thread. I am not an engineer but I already know the answer. Good luck with your experiment.

 

[svetz]

Just wanted to touch base and see how the systems been working.