Calculating inverter ratio on three strings facing different directions

[zikronix]

Trying to get estimated power production for 3 strings connected to two inverters.

14.4 / 2PW+ w/ 2 - 7.6 inverters. Phoenix, AZ

String one that faces south is 15 / 400W panels. Connected to one 7.6 inverter. So .78 ratio?
String two facing east 10 / 400W panels and string three facing west 11 / 400W panels both connected to the second 7.6 inverter. So 1.10 ratio?

The second string and third are the ones I'm trying to figure out average power production for. By basic math the ratio is 1.10 however its two strings at different azimuths with different panel counts into one inverter. Im trying to determine the power production for each string since they do face different directions. How do I do this in pv watts? Do I take one string and figure out its ratio and do the calc and then do the same with the second string thats in that inverter and just add the numbers together? Do I take each string and use calculated combined ratio of 1.10 and then add those averages together?

 

[wwhitney]

With a "nominal" DC/AC ratio of 1.1, but half facing west and half facing east, you will never have clipping. So if that's all you're worried about, you can stop now.

As for PVWatts, one approach is to run each orientation separately with a 1.0 DC/AC ratio, and dump the 8760 hourly production data for each one. Then for the inverter with two different orientations, add them up hour by hour, and then compare that to the 7.6 kW inverter rating. You'll find the sum of the AC productions never exceeds 7.6 kW.

From a vector math point of view, when the sun is shining simultaneously on both arrays (which is not always true, but when it's not true, your DC/AC ratio is less than 1), then the following applies: represent each array by a vector whose direction is the normal to the panels, and whose magnitude is the array size. Then add those vectors together, and the combined array should behave (assuming they are on separate MPPTs) like a single array represented by the vector sum, at least as far as the sun geometry factors.

So if your roof slope is 30 degrees (makes the math easy), and you have 4 kW arrays on a due east slope and a due west slope, then the vectors (in say the x-z plane) would be (-2, 4*sqrt(3)/2) and (2, 4*sqrt(3)/2). Adding them together, the array would behave like a single array of size 4*sqrt(3) = 6.9 kW pointed straight up.

Cheers, Wayne

 

[zikronix]

With a "nominal" DC/AC ratio of 1.1, but half facing west and half facing east, you will never have clipping. So if that's all you're worried about, you can stop now.

As for PVWatts, one approach is to run each orientation separately with a 1.0 DC/AC ratio, and dump the 8760 hourly production data for each one. Then for the inverter with two different orientations, add them up hour by hour, and then compare that to the 7.6 kW inverter rating. You'll find the sum of the AC productions never exceeds 7.6 kW.

From a vector math point of view, when the sun is shining simultaneously on both arrays (which is not always true, but when it's not true, your DC/AC ratio is less than 1), then the following applies: represent each array by a vector whose direction is the normal to the panels, and whose magnitude is the array size. Then add those vectors together, and the combined array should behave (assuming they are on separate MPPTs) like a single array represented by the vector sum, at least as far as the sun geometry factors.

So if your roof slope is 30 degrees (makes the math easy), and you have 4 kW arrays on a due east slope and a due west slope, then the vectors (in say the x-z plane) would be (-2, 4*sqrt(3)/2) and (2, 4*sqrt(3)/2). So adding them together, the array would behave like a single array of size 4*sqrt(3) = 6.9 kW pointed straight up.

Cheers, Wayne

That's way more info than I anticipated, yet extremely helpful and makes way more sense to me lol

 

[zikronix]

So if your roof slope is 30 degrees (makes the math easy), and you have 4 kW arrays on a due east slope and a due west slope, then the vectors (in say the x-z plane) would be (-2, 4*sqrt(3)/2) and (2, 4*sqrt(3)/2). Adding them together, the array would behave like a single array of size 4*sqrt(3) = 6.9 kW pointed straight up.

So because I suck...lol Im trying to now figure out the projected production for those two arrays. Maybe you can help further.

Slope is 23 for both azimuth is 112(east) 4000kw and 292(west) 4400kw, located in Mesa, AZ, what numbers could i put in PV to give me a rough production estimate.

 

[wwhitney]

So because I suck...lol Im trying to now figure out the projected production for those two arrays. Maybe you can help further.

Slope is 23 for both azimuth is 112(east) 4000kw and 292(west) 4400kw, located in Mesa, AZ, what numbers could i put in PV to give me a rough production estimate.

Just run each subarray separately and add up the numbers. Since you are not going to get any clipping, that should be accurate.

If you want to confirm that you aren't going to get any clipping, you can dump the 8760 hourly data points on each of the 2 subarrays above, add them up and confirm they are always less than your inverter size. Easy to do with a spreadsheet.

Cheers, Wayne

 

[zikronix]

i know im not getting clipping. ok that works thanks