Average or summation from PV ?

[mnakkach]

I want to configure 3 inverters in parallel where inverter 1 is connected to East, inverter 2 is connected to South and inverter 3 is inverter to West.

In the afternoon, if i am getting in example 1500w from PV input in inverter 1, 3000w from PV input on inverter 2 and 5000w from PV on inverter 3.

All the 3 inverters are connected in parallel, will the AC output be a combination summation of the 3 inverters and will utilize from PV 1500w+3000w+5000w ? or it will average them? or consider the lowest value?

 

[sunshine_eggo]

I must presume these are AiO inverters, in parallel, only providing a single phase and connected to the same battery.

Worth noting up front that PV input and AC power are handled by two completely independent devices, even if they're in the same box.

Your question begs another question... how do your specific inverters operate in parallel? Do they try to share the load equally? Do they cascade from one to the next? In most cases, it doesn't matter one bit as long as you have enough power for your loads.

In your situation, a total of 9500W is available from the three inverters. That can either go to loads or to battery charging.

Let's say you have 9000W of loads. How the output is handled is up to the inverters. They might split it all equally, or it might be cascaded, 5000W from the first unit, 4000W from the second and none from the third (most try to do it equally).

Their MPPTs will be pulling whatever they need to charge the battery and power the loads.

Assuming they share the output equally:

Inverter 1: pulling 1500W from PV, 1500W from battery for loads
Inverter 2: pulling 3000W from PV for loads
Inverter 3: pulling 3000W from PV for loads, 1500W from PV charging battery from inverter 1's 1500W load, 500W excess if needed for charging battery.

FWIW, you can accomplish the same kind of power distribution with three separate arrays in parallel to the same unit. In 100% of cases, a single South facing array will harvest more total energy on a given day (assuming clear skies and sunrise to sunset exposure). A split array sacrifices the total AND peak harvest to widen the output at a lower total peak.

You can simulate three separate arrays at:

PVWatts

Estimates the energy production and cost of energy of grid-connected photovoltaic (PV) energy systems throughout the world. It allows homeowners, small building owners, installers and manufacturers to easily develop estimates of the performance of potential PV installations

pvwatts.nrel.gov

You can download the hourly data and combine them in a spreadsheet. Every single time I've run the analysis, the results have been underwhelming.

 

[740GLE]

How many batteries? How many BMS?

 

[Hedges]

FWIW, you can accomplish the same kind of power distribution with three separate arrays in parallel to the same unit. In 100% of cases, a single South facing array will harvest more total energy on a given day (assuming clear skies and sunrise to sunset exposure). A split array sacrifices the total AND peak harvest to widen the output at a lower total peak.

PV of multiple orientations in parallel should result in more efficient operation of the wire between PV array and SCC and more efficient operation of SCC and inverter. Also lower temperature operation, longer life of the electronics because it operates longer at lower power instead of peaking for a shorter time.

 

[Bluedog225]

Excellent info.

Is there not a difficulty in trying to do this with one inverter? That is, too many amps from over
paneling?

 

[Hedges]

One or more SCC can be used to handle peak power from the (multiply oriented) array and peak current into battery. If over-paneled beyond that, power would be clipped.

 

[mnakkach]

I must presume these are AiO inverters, in parallel, only providing a single phase and connected to the same battery.

Worth noting up front that PV input and AC power are handled by two completely independent devices, even if they're in the same box.

Your question begs another question... how do your specific inverters operate in parallel? Do they try to share the load equally? Do they cascade from one to the next? In most cases, it doesn't matter one bit as long as you have enough power for your loads.

In your situation, a total of 9500W is available from the three inverters. That can either go to loads or to battery charging.

Let's say you have 9000W of loads. How the output is handled is up to the inverters. They might split it all equally, or it might be cascaded, 5000W from the first unit, 4000W from the second and none from the third (most try to do it equally).

Their MPPTs will be pulling whatever they need to charge the battery and power the loads.

Assuming they share the output equally:

Inverter 1: pulling 1500W from PV, 1500W from battery for loads
Inverter 2: pulling 3000W from PV for loads
Inverter 3: pulling 3000W from PV for loads, 1500W from PV charging battery from inverter 1's 1500W load, 500W excess if needed for charging battery.

FWIW, you can accomplish the same kind of power distribution with three separate arrays in parallel to the same unit. In 100% of cases, a single South facing array will harvest more total energy on a given day (assuming clear skies and sunrise to sunset exposure). A split array sacrifices the total AND peak harvest to widen the output at a lower total peak.

You can simulate three separate arrays at:

PVWatts

Estimates the energy production and cost of energy of grid-connected photovoltaic (PV) energy systems throughout the world. It allows homeowners, small building owners, installers and manufacturers to easily develop estimates of the performance of potential PV installations

pvwatts.nrel.gov

You can download the hourly data and combine them in a spreadsheet. Every single time I've run the analysis, the results have been underwhelming.

Thank you for the detailed information, but to explain what i understood in summary, in both cases if the inverter will cascade or share the load equally i won't be losing the output power from PV, so, from the input PV of 9000w, i will get AC out from the inverter as 9000W, internally between inverters, one will charge the battery and another will discharge but i will get the 9000w to the load at the end with some minimal efficiency loss due to charge/discharge i guess?

Thank for the sent link, I opened it but found it too techy to me

 

[mnakkach]

Excellent info.

Is there not a difficulty in trying to do this with one inverter? That is, too many amps from over
paneling?

How much over paneling is acceptable? BTW i am connecting a total of 7630w from PV to an inverter designed as 6000w PV max.

 

[Hedges]

Absolutely never exceed max input voltage. Compute Voc from PV as adjusted for record cold.
Don't exceed maximum short-circuit current specs. If SCC has such a spec, that is because some operating or protection mode could pull Isc from PV.
For most SCC, maximum input power is probably just guidance. Perhaps some off-brand would actually hurt themselves if more power available, but most regulate internal temperature among other parameters.

7630W STC probably rarely reaches 6000W actual.
No guarantees, but I would over-panel as many watts as I felt like or could fit. If SCC runs longer or hotter, an extra fan could help.

 

[chrisski]

I want to configure 3 inverters in parallel where inverter 1 is connected to East, inverter 2 is connected to South and inverter 3 is inverter to West.

In the afternoon, if i am getting in example 1500w from PV input in inverter 1, 3000w from PV input on inverter 2 and 5000w from PV on inverter 3.

All the 3 inverters are connected in parallel, will the AC output be a combination summation of the 3 inverters and will utilize from PV 1500w+3000w+5000w ? or it will average them? or consider the lowest value?

I tried this with a smaller system where the whole array consisted of a single inverter, and it was the sum of all three. You're asking for three inverter, but I'm not sure.

When I tried this with three strings of 300 watts of panels with one to the south, one to the East, and one to the West, the total was the sum of all three, so A + B + C=. I shut the other two lines off and measured all three one at a time, and when I turned all three on it was the sum of all three strings.

For mine, I had expected the weakest string to drag the other two strings way down because the voltage was only 2/3 of the other two, but sum of the wattages was the same.

 

[mnakkach]

Absolutely never exceed max input voltage. Compute Voc from PV as adjusted for record cold.
Don't exceed maximum short-circuit current specs. If SCC has such a spec, that is because some operating or protection mode could pull Isc from PV.
For most SCC, maximum input power is probably just guidance. Perhaps some off-brand would actually hurt themselves if more power available, but most regulate internal temperature among other parameters.

7630W STC probably rarely reaches 6000W actual.
No guarantees, but I would over-panel as many watts as I felt like or could fit. If SCC runs longer or hotter, an extra fan could help.

panel is is 49.65 voc, pmax 545w. Inverter's MPP range is 120v to 430. Max DC voltage is 450v. i am going to connect 7 in series, 2 strings parallel. Is this safe?

 

[mnakkach]

PV of multiple orientations in parallel should result in more efficient operation of the wire between PV array and SCC and more efficient operation of SCC and inverter. Also lower temperature operation, longer life of the electronics because it operates longer at lower power instead of peaking for a shorter time.

PV of multiple orientations in parallel should result in more efficient operation of the wire between PV array and SCC and more efficient operation of SCC and inverter. Also lower temperature operation, longer life of the electronics because it operates longer at lower power instead of peaking for a shorter time.

So, what makes best efficiency output? 3 inverters, each one connected to one orientation (E,S,W) or merge all orientation in parallel and limit them to 2 inverters if AC amps output is sufficient? i thought it is better to have 3 inverters so every MPPT will deal with different orientation.


Which scenario is best?

Scenario1
inverter1 South: 14 panels, 2 strings in parallel, 7 in series
Inverter2 West: 7 in series
Inveter3 East: 7 in series

Scenario2
inverter1 South: 14 panels, 2 strings in parallel, 7 in series
Inverter2 South/West: 14 panels, 2 strings in parallel, 7 to west and 7 to east

 

[sunshine_eggo]

1 MPPT per array is best in a situation where you might have severe shading and partial shading situations, and/or you want redundancy - like on a sailboat.

Worth checking this post:

How solar panels work (for dummies, volts and amps)

This post is WAY less sophisticated than one might think. This deals solely with voltage and amps and how they interact (or don't matter much) in low light conditions. There have been a handful of posts expressing concern about panels in parallel not playing well together. This is rarely a...

diysolarforum.com

Scenario 2 is more efficient.

Scenario 1 has a 3rd inverter that will have a constant idle consumption. If you're looking at cheap Chinese AiO units, count on about 40-50W per 3kW of power. If it's a 6kW inverter, it will consume as much as 2.4kWh/day, which is about what you would get on a good day from 500W of PV panels.

In other words, Scenario 1 will behave as though it has 500W less solar due to the power consumption of the 3rd inverter.

 

[mnakkach]

1 MPPT per array is best in a situation where you might have severe shading and partial shading situations, and/or you want redundancy - like on a sailboat.

Worth checking this post:

How solar panels work (for dummies, volts and amps)

This post is WAY less sophisticated than one might think. This deals solely with voltage and amps and how they interact (or don't matter much) in low light conditions. There have been a handful of posts expressing concern about panels in parallel not playing well together. This is rarely a...

diysolarforum.com

Scenario 2 is more efficient.

Scenario 1 has a 3rd inverter that will have a constant idle consumption. If you're looking at cheap Chinese AiO units, count on about 40-50W per 3kW of power. If it's a 6kW inverter, it will consume as much as 2.4kWh/day, which is about what you would get on a good day from 500W of PV panels.

In other words, Scenario 1 will behave as though it has 500W less solar due to the power consumption of the 3rd inverter.

Ok, convinced with scenario 2, but panel is is 49.65 voc, pmax 545w. Inverter's MPP range is 120v to 430. Max DC voltage is 450v. i am going to connect 7 in series, 2 strings parallel. Total will 7630 w from PV per inverter. It is a Voltronic V IV, is this ok or too much over paneling ?

 

[sunshine_eggo]

If you stay under PV input voltage and current limits, over panel as much as you like.

I would consider 8S. Should be good down to about -10°C.

IIRC, similar units have around an 18A PV input current limit. There are mixed reports on how far you can go, but most agree that you can go slightly over that - maybe 20-22A.

 

[mnakkach]

If you stay under PV input voltage and current limits, over panel as much as you like.

I would consider 8S. Should be good down to about -10°C.

IIRC, similar units have around an 18A PV input current limit. There are mixed reports on how far you can go, but most agree that you can go slightly over that - maybe 20-22A.

To join east (7) and west (7) on one string, do i need any device to block the return of current from 1 string to the other?

 

[sunshine_eggo]

To join east (7) and west (7) on one string, do i need any device to block the return of current from 1 string to the other?

Only if you plan to completely cover one array or the other and prevent any light whatsoever hitting those panels. As long as they get ambient light, even shaded, they will have sufficient voltage to prevent the other half of the array from backfeeding current through it.

 

[sunshine_eggo]

Thank for the sent link, I opened it but found it too techy to me

If this is too "techy" for you, you should consider if you're capable of this.

You simply need to enter your exact location, your array size and orientation. On the output page, you can download an hourly .CSV file. You would repeat that for the 3 arrays and then manually combine the CSV data into a single spreadsheet.

It MIGHT take you an hour to do if you take a couple breaks, make mistakes and end up doing it over 2-3 times. An hour's worth of planning on a project of this type always pays dividends.

 

[Ampster]

To join east (7) and west (7) on one string, do i need any device to block the return of current from 1 string to the other?

It would be better to put them on different MPPT controllers. Others with more experience can give you specifics with more details.
I do agree with the answer from @sunshine_eggo that return current should not be an issue. My gut feel is that is would not be optimum to keep them on the same MPPT controller.

 

[mnakkach]

Only if you plan to completely cover one array or the other and prevent any light whatsoever hitting those panels. As long as they get ambient light, even shaded, they will have sufficient voltage to prevent the other half of the array from backfeeding current through it.

Does it make it better in scenario 3? this may distribute the PV load equally on the 2 inverters?

Scenario3
inverter1 West/South: 14 panels, 2 strings in parallel, 7 to West and 7 to South
Inverter2 East/South: 14 panels, 2 strings in parallel, 7 to East and 7 to South

Scenario2 was
Scenario2
inverter1 South: 14 panels, 2 strings in parallel, 7 in series (ALL to South)
Inverter2 East/West: 14 panels, 2 strings in parallel, 7 to West and 7 to East