One inverter with 20 strings All string facing one orientation or 10 strings oposite?

[bendesa]

Hi All,

Imagine a roof with the same tilt and one roof faces east and the other roof faces west. The two roofs are exactly the same.

We have inverters to which 20 strings can be connected. Do we install 20 strings in 1 direction, for example towards the west or
do we divide the number of strings over the two roofs.

So 10 strings to the east and 10 strings to the west

What do you guy think?

Kind regards

Ben

 

[GoldDigger]

If you were dealing with an off grid system, charging batteries, the extended charging time that comes from splitting orientation may be a large benefit.
If you have a grid-interactive system without net metering, the better average to peak ratio will favor split orientations.
If your design overpanels your inverter (peak panel output greater than inverter input rating) then split orientation will waste less energy.
If you have grid-interactive, with TimeOfDay pricing, see which single orientation will get you the best price for your output.
Finally, your local weather conditions may give you greater sun exposure over the year with one orientation of the other (cloud cover in morning versus evening, for example). Use simulation software to check.

 

[LarryFine]

The half-and-half layout makes the most sense to me, especially if the exposure is genuinely equal through the year.

The peak output may be slightly lower, but the average might be higher. Unless I don't know what I'm talking about.

 

[retirede]

The half-and-half layout makes the most sense to me, especially if the exposure is genuinely equal through the year.

The peak output may be slightly lower, but the average might be higher. Unless I don't know what I'm talking about.

We should all include that last sentence qualifier on all of our posts!

 

[bendesa]

Hi All,

Same question but not facing West & East but North & South

Imagine a roof with the same tilt and one roof faces north and the other roof faces south. The two roofs are exactly the same.

No shading and it's a Grid Tied PV System

We have inverters to which 20 strings can be connected. Do we install 20 strings in 1 direction, for example towards the north or towards the south? Or do we divide the number of strings over the two roofs.

So 10 strings to the north and 10 strings to the south

What do you guy think?

Kind regards

Ben

 

[ggunn]

Hi All,

Same question but not facing West & East but North & South

Imagine a roof with the same tilt and one roof faces north and the other roof faces south. The two roofs are exactly the same.

No shading and it's a Grid Tied PV System

We have inverters to which 20 strings can be connected. Do we install 20 strings in 1 direction, for example towards the north or towards the south? Or do we divide the number of strings over the two roofs.

So 10 strings to the north and 10 strings to the south

What do you guy think?

Kind regards

Ben

It depends. For maximizing production the modules should all face south, or as as many of them as possible. If there is not room to face them all south the tilt angle must be considered, and if it is significant you 'll need to determine whether the expense of the north facing modules is worth installing them at all. If it is a string inverter without optimizers no string should contain modules facing in both orientations.

 

[solarken]

Hi All,

Same question but not facing West & East but North & South

Imagine a roof with the same tilt and one roof faces north and the other roof faces south. The two roofs are exactly the same.

No shading and it's a Grid Tied PV System

We have inverters to which 20 strings can be connected. Do we install 20 strings in 1 direction, for example towards the north or towards the south? Or do we divide the number of strings over the two roofs.

So 10 strings to the north and 10 strings to the south

What do you guy think?

Kind regards

Ben

You might want to go to https://pvwatts.nrel.gov/ and enter the PV system info and location info and run a simulation to show the production for the different orientation options. I think you will find that unless the slope is very low, like 4 degrees or so, that installing on the north facing slope will result in significantly lower production in most areas of the US. If I have East and West I usually install on both rather than favor one, because as others have said the peaks are ower, the energy is spread over more of the day, and you may be able to install lower inverter capacity or at least minimize clipping.

 

[wwhitney]

So, I found the graph below from a random website, https://solarlove.org/solar-cell-model-and-its-characteristics/

If it's accurate, you can see the effect of putting two otherwise identical but differently illuminated strings in parallel, which forces them to be at the same voltage. If all the peaks were aligned over a single voltage, then there'd be zero loss from combining the strings. But I see V_mpp varying from about 0.45V per call at 20% illumination to maybe 0.51V at 100% illumination.

So if you put two of those strings in parallel, the MPPT algorithm will pick a voltage somewhere in between those values, where the combined power is maximized. That will be the voltage where the two curves have equal and opposite slopes, so that their sum has zero slope. Maybe for the example it's 0.47V per cell?

And then you for each string you can compare the power at 0.47V to the power at its V_mpp. So maybe the 20% illuminated string is making 0.32W per cell instead of 0.35W per cell, and maybe the 100% illuminated string is making 2.10W per cell instead of 2.17W per cell. [It's hard to read the graphs accurately.] That would mean you are getting 2.45W per pair of cells, rather than 2.52W per pair of cells if the strings were on separate MPPT inputs, for a loss of about 3% for the case of 100% illumination and 20% illumination in parallel.

Of course, it would be best to look at curves for 5% and 10% illumination as well.

Cheers, Wayne

 

[wwhitney]

For maximizing production the modules should all face south, or as as many of them as possible.

That wasn't the question. The question was, if all your roof planes are covered with PV, and you have already split the panels up into strings, where each string is only all north facing or all south facing, what are the pros and cons of keeping each MPPT input either all north or all south, vs paralleling a north string and a south string on one MPPT input?

For the case of East and West, there's some inverter capacity/cost savings to be had by putting one East string and one West string together at the inverter. Each string will produce peak power at a different time of day, so in term of peak power, (East + West together) < (East alone) + (West alone).

[I'm not familiar enough with string inverters to know the following: if a 7.6 kW inverter (say) has two different MPPT inputs, is each one rated for only 3.8 kW? If so, you'd have to put an east string and a west string in parallel on a single input in order to use the full inverter capacity, and you'd have some losses as described in my previous post. But if each MPPT input is individually rated, say, 5.0 kW, with 7.6 kW a joint limit, then you could separate the East and West strings on separate inputs and have no losses from string orientation mixing.]

For the case of due North and due South, however, peak power from each string will occur at the same time of day for each string. So there's no inverter capacity savings to be had by mixing North and South together, and you might as well keep them segregated at the inverter inputs.

Cheers, Wayne

 

[wwhitney]

That would mean you are getting 2.45W per pair of cells, rather than 2.52W per pair of cells if the strings were on separate MPPT inputs, for a loss of about 3% for the case of 100% illumination and 20% illumination in parallel.

That should be 2.42W vs 2.52W, or a 4% loss.

Cheers, Wayne

 

[ggunn]

That wasn't the question. The question was, if all your roof planes are covered with PV, and you have already split the panels up into strings, where each string is only all north facing or all south facing, what are the pros and cons of keeping each MPPT input either all north or all south, vs paralleling a north string and a south string on one MPPT input?

There's not a whole lot of difference. Voltage doesn't change much with orientation, so although the MPPT point will be different for a north facing vs a south facing array at any point in time and the MPP tracker will pick a point between them, it won't be that much different from what a single facing array would be. For dual tilt racking on flat roofs it isn't unusual to put a string facing one way with another facing the other way in parallel on the same MPPT input. Of course, the more extreme the difference in tilt, the more the difference in voltage will be, so it's a judgement call.

 

[wwhitney]

There's not a whole lot of difference.

Agree (post #8) but for N and S there's no upside, unlike for E and W (post #9).

[Do you know the answer to the parenthetical question in post #9?]

Cheers, Wayne

 

[ggunn]

Agree (post #8) but for N and S there's no upside, unlike for E and W (post #9).

[Do you know the answer to the parenthetical question in post #9?]

Cheers, Wayne

Sorry; I do not quite understand the question. Without using optimizers (a la SolarEdge) you can put parallel strings of different orientations on the same MPPT without very much loss due to tracking point compromise as long as you do not mix orientations within a string. With optimizers you can mix orientations, shading, even module manufacturers and power ratings within strings.

 

[wwhitney]

Sorry; I do not quite understand the question.

Well, I was looking at the spec sheet of Sol-Ark 12K-2P (a hybrid inverter) recently, and I noticed that it says

Max PV Power = 12,000W
Max DC Voltage = 450V @ 20A, 500V @ 18A
MPPT Voltage Range = 150V-425V
Number of MPPT = 2
Max DC current per MPPT = 20A.

So it looks like for that model, you could put 400V @ 20A on one MPPT input (8 kW), and then 400V @ 10A on another MPPT input (4 kW). I.e. if you have an equal number of North facing strings and South facing strings, if the numbers worked out you could put all the South facing strings on one input, and all the North facing strings on another input.

Whereas if the max DC current per MPPT were 15A, that's only 6kW at 400V, and you'd need to split the available DC power more evenly between the two inputs. Which would likely mean splitting both North strings and South string evenly between the 2 inputs.

So how common are specs like the above, vs a hypothetical string inverter where each MPPT input has a power limit near (total power / # of inputs)?

Cheers, Wayne