diy solar

diy solar

Can Growatt Inverter do this?

I’m interested in any UL-Listed GTILs that can be connected in parallel
I think a search for GT Inverters will get you some results. You will have to search within that general classification for UL Approved and ones that can be limited to no export. A search for GTIL only comes up with posts on this forum so I suspect GTIL is not a commonly used term.
What are you using to control your Enphase Micros and does it need to be wired into the critical path like a hybrid inverter or can it connect on parallel through it’s own circuit like a GTIL
The Enphase Envoy with the addition of CTs for consumption is where the programming is done. They can wired to any circuit breaker on the main panel or a sub panel. The only requirement is that the bus bar is large enough so that the NEC 120% rule is followed. In my case I put them on a sub panel that will be fed by my soon to be installed hybrid. That way they can be AC coupled to the hybrid and I can use their production when the grid is down. The hybrid will also have some DC coupled solar panels on a west facing direction. The hybrid will be programmed to not export more than is allowed in my PTO plus the 1kW extra allowed. The hybrid will be installed to code but no permission will be requested from PGE. The total kW capacity will be 2kW more than my PTO but the export limit will keep me within my contractual arrangement with PGE.
 
I think a search for GT Inverters will get you some results. You will have to search within that general classification for UL Approved and ones that can be limited to no export. A search for GTIL only comes up with posts on this forum so I suspect GTIL is not a commonly used term.
A search for no export generally gets you a large number of ‘in series’ hybrid inverters and no clamp-sensor-based GT inverters that can be connected in parallel (the way a Microinverter-based string gets connected into a mains panel).


The Enphase Envoy with the addition of CTs for consumption is where the programming is done. They can wired to any circuit breaker on the main panel or a sub panel. The only requirement is that the bus bar is large enough so that the NEC 120% rule is followed.

So the envoy connects similar to the way a Microinverter-based PV string connects to the main panel?

That’s what I’m looking for.

I’ve already got my Microinverter-based string connected into my main panel and pretty much ‘using up’ the 120% rule so it sounds as though I may have an issue connecting an Envoy in addition. Is there a way to connect a Microinverter-based string ‘through’ the Envoy?

In my case I put them on a sub panel that will be fed by my soon to be installed hybrid. That way they can be AC coupled to the hybrid and I can use their production when the grid is down. The hybrid will also have some DC coupled solar panels on a west facing direction.

Yes, I understand that with a hybrid in the power path, all sorts of great things become possible. I’ve got only 3” of conduit between my main breaker on the outside of the house and my main panel on the other side of the same wall, so wiring a hybrid inverter into the main power path essentially means adding an entirely new main panel as a subpanel and moving all breakers over (prohibitively expensive).

The hybrid will be programmed to not export more than is allowed in my PTO plus the 1kW extra allowed. The hybrid will be installed to code but no permission will be requested from PGE. The total kW capacity will be 2kW more than my PTO but the export limit will keep me within my contractual arrangement with PGE.
You’ll probably be fine but you should probably be aware that your NEM agreement specifically requires review and permission by PG&E before making any changes and certainly before making use of your 1kW addition allowance.

I was planning to add a hybrid and another 1kW of panels precisely for that reason and PG&E explicitly told me that I would void my agreement if not getting their approval before adding my 1kW of additional solar generation.

As long as you are keeping your export under whatever max kW is in your current agreement, you should be fine, but if you limit to 1kW beyond that cap, you are taking a risk that they find out and tell you you’ve voilated the terms of your NEM agreement.

Since my discussions with PG&E, I’ve changed course and am now only interested in preserving as much of my already-authorized generation credits by using some additional DC-coupled PV power to offset as much self-consumption as possible (especially during peak hours).
 
So the envoy connects similar to the way a Microinverter-based PV string connects to the main panel?
No, the Envoy is powered by a separate breaker which can be as small as 10Amps. I use a 15 Amp dual pole breaker which also powers my Emporia Energy meter. It uses a powerline protocol to communicate with the IQ7s.
A search for no export generally gets you a large number of ‘in series’ hybrid inverters and no clamp-sensor-based GT inverters that can be connected in parallel (the way a Microinverter-based string gets connected into a mains panel).
As I mentioned I would start the search for all GT inverters to get the largest list. Then it may be a process of manually drilling down into the details of each inverter on that list to find which ones can be export limited. Off hand I only know for a fact that the Envoy can be programmed but thought the Solaredge could also. There is a market for that capability but maybe it exists more frequently in hybrids.
You’ll probably be fine but you should probably be aware that your NEM agreement specifically requires review and permission by PG&E before making any changes and certainly before making use of your 1kW addition allowance.
My interpretation is that the wording is vague. Did the approve your Growat? My risk profile is I can handle a contractual dispute but I do not want to take a chance with non UL approved equipment and a non code compliant installation.
I was planning to add a hybrid and another 1kW of panels precisely for that reason and PG&E explicitly told me that I would void my agreement if not getting their approval before adding my 1kW of additional solar generation.
Of course they would tell you that. I will take my chances. My NEM PTO has little value now that they have screwed with the hours and added NBCs and taken away my really good EV rate. The value to me of all the work to add a hybrid inverter and subpanels is I can self consume much of my solar and just use the grid as backup. I mostly charge my EVs during the day from solar anyway. I should never have to purchase any power at peak periods, except in unusual circumstances.
 
Last edited:
No, the Envoy is powered by a separate breaker which can be as small as 10Amps. I use a 15 Amp dual pole breaker which also powers my Emporia Energy meter. It uses a powerline protocol to communicate with the IQ7s.
I understand that the Envoy controls the Enphase Microinverters using power line communication, but it sounds like is connected in parallel, as I had hoped. So the string of Enphase inverters comes into the main panel through it’s own breaker and the Envoy comes into the main panel through it’s own (independent breaker)? No direct electrical connection between Envoy and microinverters (power line communication only)?

And that means the Envoy performs 3 functions:

1/ throttle-back Enphase Microinverters yo avoid export when there is too much generation.

2/ generate AC power from battery to supply loads when programmed to do so.

3/ consume AC power to charge battery when programmed to do so (ie: off-peak and/or to avoid export).

Is that about right?

As I mentioned I would start the search for all GT inverters to get the largest list. Then it may be a process of manually drilling down into the details of each inverter on that list to find which ones can be export limited.

It’s export limit I’m seeking - most hybrid inverters will limit export - it’s use of a clamp sensor to limit export (which allows for parallel connection).

A hybrid inverter in the power path (serial connection) can sense consumption and/or export because the power is flowing through it.

A GTIL uses a clamp sensor to sense consumption and/or export and do does not need to be wired into the current path (hence allowing it to be wired in through a parallel connection).

Off hand I only know for a fact that the Envoy can be programmed but thought the Solaredge could also. There is a market for that capability but maybe it exists more frequently in hybrids.

My interpretation is that the wording is vague. Did the approve your Growat? My risk profile is I can handle a contractual dispute but I do not want to take a chance with non UL approved equipment and a non code compliant installation.
The risk you are taking is that you’ll be subject to their interpretation should they decide that you’ve violated the terms of your NEM agreement and inform you that it is now void. You’ll just start getting billed.
Of course they would tell you that. I will take my chances. My NEM PTO has little value now that they have screwed with the hours and added NBCs and taken away my really good EV rate.

You’re preaching to the choir. I installed 120% of consumption in 2016 and got a true-up bill for the first time this August (for over a third of my old bill).

By 2024, when we all get moved to 4-9 Everyday, my annual true-up will probably be up over 50% of what I used to pay previously before I added solar…

I got onto NEM1 just before they closed it so it’s still worth something over NEM2 (because import is not taxed as it is under NEM2, only net import at True Up is taxed).
The value to me of all the work to add a hybrid inverter and subpanels is I can self consume much of my solar and just use the grid as backup. I mostly charge my EVs during the day from solar anyway. I should never have to purchase any power at peak periods, except in unusual circumstances.

That’s the way to do it. I’m preparing to do the same thing, but not until we get our first EV in ~2023.

If you can charge your EV while the sun is shining, that’s perfect and you can keep grid consumption down to peak demand and what you need to close the gaps during an extended period of bad weather.

I’m hoping to essentially preserve NEM credits to allow me to charge an EV whenever needed (typically after work hours).
 
No direct electrical connection between Envoy and microinverters (power line communication only)?
No. by definition a parallel connection is a direct electrical connection. The Envoy sees the same voltage and frequency that the whole house sees other than voltage drop due to wire resistance.
And that means the Envoy performs 3 functions:

1/ throttle-back Enphase Microinverters yo avoid export when there is too much generation.
Yes
2/ generate AC power from battery to supply loads when programmed to do so.
If you are talking about the Enphase battery, I don't know the exact process except that the Enphase battery is AC coupled. That implies that the Enphase battery uses separate microinverters DC coupled to the actual battery pack.
3/ consume AC power to charge battery when programmed to do so (ie: off-peak and/or to avoid export).

Is that about right?
Again that is internal to the Enphase battery and is not how my hybrid inverter controls the Enphase micros. That control by my hybrid is strictly through the AC coupling by frequency shift and possibly volta

A hybrid inverter in the power path (serial connection) can sense consumption and/or export because the power is flowing through it.
Yes, but many hybrids also can use an external CT so it can react to the power flowing through the external CT.
A GTIL uses a clamp sensor to sense consumption and/or export and do does not need to be wired into the current path (hence allowing it to be wired in through a parallel connection).
An external sensor can be place anywhere there is current. The reason I want to add an external CT to my hybrid is that I want to measure the current close to the meter. That is where the export is measured for PGE.
 
The risk you are taking is that you’ll be subject to their interpretation should they decide that you’ve violated the terms of your NEM agreement and inform you that it is now void. You’ll just start getting billed.
Most of my career involved some form of risk assessment. My assessment is that PGE is not very likely to to discover that I may be violating any terms of their contract. Other than SGIP recipients I am not sure they they even have a process to monitor production of NEM agreements. In a former location I purposely set my hybrid for a higher setting for the last three months and never heard anything. Even if they did my strategy is to apologize and promise to never do it again. The concept is it is much easier to ask forgiveness than ask permission. The worst case if all that fails, is that I lose NEM 2.0. If I were on NEM 1.0 it would be a bigger downside. However if I installed non UL equipment and did not follow code the downside is much greater to me. I am not suggesting that everyone should follow my path. Risk management is a process of accepting the risks that are tolerable and mitigating the risks that are intolerable.
 
Last edited:
Most of my career involved some form of risk assessment. My assessment is that PGE is not very likely to to discover that I may be violating any terms of their contract. The downside is that I lose NEM 2.0. If I were on NEM 1.0 it would be a bigger downside. However if I installed non UL equipment and did not follow code the downside is much greater. I am not suggesting that everyone should follow my path. Risk management is a process of accepting the risks that are tolerable and mitigating the risks that are intolerable.
If I was on NEM 2.0, I’d probably approach things as you are.

And certainly agree that the cost of losing a grandfathered NEM agreement is nothing compared to the cost of having an accident because equipment was not UL certified.


For the small experimental system I’ve got cobbled together now, my assessment is that the risk associated with my DIY 560Ah LiFePO4 battery is at least 10 times greater than the risk associated with the fact that the 2 small 1kW GTIL inverters I’m using are not UL certified…

(They being said, I’ll be happy if my ‘permanent build’ in 2023 can be based on UL-certified boxes ;)).
 
No. by definition a parallel connection is a direct electrical connection. The Envoy sees the same voltage and frequency that the whole house sees other than voltage drop due to wire resistance.

Yes

If you are talking about the Enphase battery, I don't know the exact process except that the Enphase battery is AC coupled. That implies that the Enphase battery uses separate microinverters DC coupled to the actual battery pack.

Again that is internal to the Enphase battery and is not how my hybrid inverter controls the Enphase micros. That control by my hybrid is strictly through the AC coupling by frequency shift and possibly volta
So is it the Enphase battery that communicates and controls the Enphase Microinverters or is it a separate Hybrid Microinverter?

I’m not understanding whether the Enphase Battery in only a slave (to an appropriately-enabled Microinverter) or whether is is or can be the master…

Yes, but many hybrids also can use an external CT so it can react to the power flowing through the external CT.

An external sensor can be place anywhere there is current. The reason I want to add an external CT to my hybrid is that I want to measure the current close to the meter. That is where the export is measured for PGE.
Any inverter that supports an external CT sensor is of interest to me, but if the inverter needs to be positioned between the grid and the load, that’s a lot less exciting (at least for my requirements).
 
So is it the Enphase battery that communicates and controls the Enphase Microinverters or is it a separate Hybrid Microinverter?
I don't know much about the Enphase battery. What I do know is it is only available through installers and it is expensive on a dollars per kWh metric. I saw one on a display at the local CED Greentech. It had two or three micros and in slots and that was all I could identify. That is why I assumed it would be AC coupled to the rooftop micros.
I’m not understanding whether the Enphase Battery in only a slave (to an appropriately-enabled Microinverter) or whether is is or can be the master…
I have never thought of a master slave relationship but in a typical AC coupled system the controlling inverter could be the referred to as the master.
In the same context any GT inverter is a slave to the grid in the sense that the grid controls whether the micro is on or off or on some cases has reduced its output in response to frequency or voltage. I have never seen the term master or slave used in describing AC coupled arrangements
Any inverter that supports an external CT sensor is of interest to me, but if the inverter needs to be positioned between the grid and the load, that’s a lot less exciting (at least for my requirements).
I understand your reluctance to do the wiring necessary but there are limits on functionality with that approach. Again, I am not trying to convince you to change your approach. However for those readers that want that functionality there are a lot of benefits once they cross that bridge. That is particularly true for backup scenerios and load shifting scenerios.
 
Last edited:
I don't know much about the Enphase battery. What I do know is it is only available through installers and it is expensive on a dollars per kWh metric. I saw one on a display at the local CED Greentech. It had two or three micros and in slots and that was all I could identify. That is why I assumed it would be AC coupled to the rooftop micros.
Oh, I must have misunderstood - I thought you owned one. So what are you using to throttle-back output of your Enphase Microinverters and what battery are you using in your system?
I have never thought of a master slave relationship but in a typical AC coupled system the controlling inverter could be the referred to as the master.
In the same context any GT inverter is a slave to the grid in the sense that the grid controls whether the micro is on or off or on some cases has reduced its output in response to frequency or voltage. I have never seen the term master or slave used in describing AC coupled arrangements
On and off control is very different than throttling control. I looked into the Enphase solution at one point but I can’t recall whether the Enphase battery can communicate with and control the Enphase Microinverters or whether the Enphase solution requires a seperate Enphase Hybrid Inverter to do so…

I understand your reluctance to do the wiring necessary but there are limits on functionality with that approach. Again, I am not trying to convince you to change your approach. However for those readers that want that functionality there are a lot of benefits once they cross that bridge. That is particularly true for backup scenerios and load shifting scenerios.
Absolutely - not installing an ‘in-the-current-path’ hybrid inverter means you lose any capability for automatic backup. That is the lowest priority for me (happy to handle rarely-used backup manually by throwing a few switches when needed).

My top priority is self-consumption with time-shift capability.

I’m not certain what you meant by ‘load shifting’ but if you meant time-shifting, the GTILs are close to ideal for that.

I literally started using mine by controlling
them with a pair of lamp timers. Use an SCC to charge up a battery during the day and then turn on the GTILs to cover self-consumption during peak hours in the evening.

I’ve since discovered that my SCC includes a pair of programmable dry contacts that can be programmed to close both via an internal timer as well as sensed battery voltage (a crude estimate of battery SOC),

This had allowed me to add to the sophistication of my time-shift system so that I can now ‘reserve’ a sufficient amount of battery capacity to be used during peak hours while also setting an ‘it’s a high-production-day-with-plenty-of-solar-generation’ SOC level which allows me to start self-consumption before peak hours so no potential energy generating gets wasted (because the battery got filled before peak period begins).

Since I started using my system a month ago, my SCC has never been out of boost/bulk charging mode and I don’t believe I’ve wasted a single Wh of energy potential generated by my new DC-coupled array.

I just checked today and net energy consumed from mid-September to mid-October has dropped by over 95% versus the same period a year ago (pretty happy).
 
So what are you using to throttle-back output of your Enphase Microinverters and what battery are you using in your system?
It is in my signature. I use an Outback Skybox and 48 EVE 280 Ahr cells.

I’m not certain what you meant by ‘load shifting’ but if you meant time-shifting, the GTILs are close to ideal for that.
I am not familiar with the term time shifting but maybe it is the same thing. I use the term load shifting to describe two things. in the first case I shift my loads like EVs, water heaters and other loads to off peak. Secondly, I cover the other house loads like lights and refrigerator with power from the inverter. To me that is what self consumption implies.
 
It is in my signature. I use an Outback Skybox and 48 EVE 280 Ahr cells.
Not familiar with Skybox - does it provide the needed control to the throttle-back your Enchase Microinverters?
I am not familiar with the term time shifting but maybe it is the same thing. I use the term load shifting to describe two things. in the first case I shift my loads like EVs, water heaters and other loads to off peak. Secondly, I cover the other house loads like lights and refrigerator with power from the inverter. To me that is what self consumption implies.
Yes, charging your EV or heating your water after peak hours instead of in the middle of peak hours sounds like ‘load shifting’.

But I’m pretty sure that running loads during peak period off of energy stored in a battery using an inverter is ‘time shifting’ (shifting the useful time of your solar generation to later in the day after the sun has gone down).
 
Not familiar with Skybox - does it provide the needed control to the throttle-back your Enchase Microinverters?
Yes, the Skybox was designed to be grid interactive. It is an all in one system. It AC couples to my Enphase inverters. That is only needed when the grid is down and loads and battery charging are less than the micros produce.. The rest of the time the micros full output goes to the loads then the grid.
 
Yes, the Skybox was designed to be grid interactive. It is an all in one system. It AC couples to my Enphase inverters. That is only needed when the grid is down and loads and battery charging are less than the micros produce.. The rest of the time the micros full output goes to the loads then the grid.
That’s nice Technology.

Do you know whether the Skybox is using an Enphase-specific PLC protocol to communicate with the Enphase Microinverters or one of the standard protocols such as ‘Frequency-shift’?

My 4kW Microinverters-based grid-tie array was established just before the new class of ‘smart’ Microinverters supporting throttling protocols such as frequency shift emerged.

If my NEP Microinverters ever give up the ghost, I’ll definitely be replacing them with smarter Microinverters.

But in the meantime, running the full array at max and shutting the whole array down are the only controls I have.

A bigger batttery + a 4kW AC charger would allow me to run the array once per day to fill the battery but off-grid / backup is not my priority.

One compensation about missing NEM1 is you got superior Technology ;).
 
Do you know whether the Skybox is using an Enphase-specific PLC protocol to communicate with the Enphase Microinverters or one of the standard protocols such as ‘Frequency-shift’?
Both the micros and the Skybox are 1741SA and CA Rule 21 compliant. I think that includes frequency and voltage but I rely on the devices to figure out the details. As far as I know the Enphase PLC protocol is proprietary and they are the only ones that use it.
 
Both the micros and the Skybox are 1741SA and CA Rule 21 compliant. I think that includes frequency and voltage but I rely on the devices to figure out the details. As far as I know the Enphase PLC protocol is proprietary and they are the only ones that use it.
That makes sense. I’ve been tracking the CA Rule 21 solutions and last time I checked (~1 year ago), the proprietary PLC- based controls were working more reliably than frequency-shift-based controls (at least when mixing and matching between vendors).

So if you are satisfied with the way in which your Skybox is throttling back your CA Rule 21-compliant Enphase Microinverters, that’s good progress…
 
the proprietary PLC- based controls were working more reliably than frequency-shift-based controls (at least when mixing and matching between vendors).
I am not sure I understand? How would a proprietary PLC control even work with another vendors product? The goal of standards like 1741 and Rule 21 are to have greater interoperability between products.
 
I am not sure I understand? How would a proprietary PLC control even work with another vendors product?

They wouldn’t (hence the ‘proprietary’ part…).
The goal of standards like 1741 and Rule 21 are to have greater interoperability between products.
Absolutely. But they are not based on PLC.

1741 and Rule 21 built upon the standard ‘grid present’ specification for all grid-tie inverters but especially Microinverters.

Grid frequency must be within a standard min-max range and grid voltage must also be within a standard min-max range. If the grid signal seen by the grid-tied inverter is out of this range (for either frequency or voltage), the inverter must shut down.

This was for safety and this is the level of technology contained in my pre-1741, pre-Rule 21 Microinverters.

The 1741, Rule 21 standard built upon that by specifying that output power must reduce in steps from maximum output at maximum allowable frequency to no output (off) at minimum allowable frequency (and please don’t quote me, I may have it backwards, but it is one or the other).

And the same for voltage. I’m not sure whether compliant inverters must support both throttling methods or only one.

To make this work, there must be a hybrid inverter in between the grid and the inverter(s) to control/modify the grid frequency and/or voltage.

This is not Power Line Communication - this is updating a crude binary communication protocol (ON or OFF) to a continuous several-bits of greyscale protocol by requiring finer accuracy in both sensing and control of grid signal (frequency and/or voltage).

The ‘smart grid’ of the future will be based on this type of protocol - the grid itself will modify frequency when it wants less solar power to be exported (from all generators).

Power Line Communucation is a much more sophisticated protocol that involves sending packets of addressed digital information to individual receivers on the home’s power line.

PLC is used for power line Ethernet but each of those PLC Ethernet solutions is proprietary (they cannot be mixed-and-matched). Same for Microinverter monitoring gateways - they are based on PLC. My NEP gateway plugged into any 120V socket in the home is able to communicate with all of my 240VAC NEP Microinverters to download status and output data.

But each of those gateways is based on a proprietary PLC protocol. An NEP gateway will not work with Enphase Microinverters and Vica-versa.

Communication digitally using PLC is far, far more robust and reliable than communicating in analog using frequency or amplitude as the analog control variable.

While the 1741 / CA Rule 21 standards should allow interoperability, a year ago getting that interoperability to work correctly was proving to be much more problematic than expected.

Victron and AP Systems put out a press release to brag about being one of the first combos to prove interoperability: https://emea.apsystems.com/wp-content/uploads/2018/03/APsystems_YC500_Victron-MultiPlus_2018_WP.pdf

When I spoke with Magnum regarding their PAE hybrid inverter supporting 1741 / CA Rule 21 one year ago, they essentially scared me away and told me that analog throttling of a DC coupled array using string voltage under control of an SCC was orders of magnitude more reliable than attempting to throttle AC-coupled Microinverters using frequency or amplitude shift.

I mean, think about it: every load in your house consuming energy monkeys with the grid signal (both frequency and voltage). Motors (ie: fridges and washing machines / dryers) are especially bad for this.

The new standard for Rapid Disconnect is a first effort to standardize PLC-like communication (digital heartbeat signal).

Building upon that, I suspect we will eventually see standardized PLC communication for Microinverter throttling, but that will take years, and until then, staying within a ‘walled garden’ which relies on proprietary PLC communication is likely to be much more effective than attempting to mix and match between vendors interoperating trough frequency and/or voltage shift.

At least that was the situation one year ago - perhaps everything is golden today.

The way my vendor, NEP, is solving this 1741 / CA Rule compatibility issue is that in addition to supporting frequency and voltage shift standards, they have instituted a proprietary PLC-based protocol to throttle-back microinverter output under the digital control of their gateway.

With the gateway as the master monitoring what level of power is needed and communicating reliably to the Microinverters using PLC, the possibility of ‘thrashing’ or analog throttling resulting in harsh ON/OFF control is greatly reduced.

I’m pretty certain Enphase has a solution based on exactly this. Whether it is the Enphase battery and/or an Enphase Hybrid Inverter, they communicate with the Enphase Microinverters using a proprietary Enphase PLC protocol to throttle output.

The inverters also need to support analog throttling through frequency and/or voltage shift to be 1741 / CA Rule 21 compliment - it is just that using the proprietary PLC protocol (when possible) will result in superior performance (and specifically lower likelihood of thrashing).
 
Last edited:
I know AC coupling is complicated. My analysis is it was easier to implement 1741 and Rule 21 than to implement features in an inverter to be able control another inverter. For example, I bought my Skybox in early 2018 and it took Outback another nine months to get AC coupling implemented in firmware. I rarely needed it so it was not an issue for me then. I understand some of the other manufacturers are still wrestling with AC coupling.
 
The inverters also need to support analog throttling through frequency and/or voltage shift to be 1741 / CA Rule 21 compliment - it is just that using the proprietary PLC protocol (when possible) will result in superior performance (and specifically lower likelihood of thrashing).
Can anyone point me to a document that says enphase is using their PLC to control the inverter output similar to what 1741/CA21 requires? What little I have seen it looks like the enphase only uses the PLC for data gathering, not control.

BTW: I agree that something like the enphase PLC could be a better way to control AC coupled inverters.......
 
Back
Top