Parallel vs series ; two classics vs 1; different array angle(s) ; shading

[ebenbayer]

So this question seems like it should be answerable by searching but I've read a few (a lot) of threads and still am not clear. So i'm looking for some advice on strategy for wiring an expanded array (parallel, series, etc):

I have a classic 250 w/ 5 panels in series, (305 W panels). ~225 VOC mounted on a 45 degree roof w/ no shading ~9 am to 4 PM

I'm planning to add more panels to the system... and I happen to have another classic 200 which is currently not utilized. Here are the option's I'm considering:

1) 5 more panels in parallel wired into the same classic 250.

The second set of panels will be in the same plane as the current panels, but will be mounted past the end of the roof eve's on a ground based array, I was thinking of mounting them much closer to Vertical to shed winter snow.

QUESTION: Can I parallel strings at different angles w/ good results or will this notably decrease my array efficiency?

Issue 2: Because these panels are mounted closer to the ground I expect they will encounter more shading than the roof array

QUESTION: I understand parallel strings are more shade tolerant, but given I have another CC around would it be worthwhile to employ this for the second set of panels ? (in which case I would go to 6 panels and do two strings of 3 panels wired to the classic 200 at the different view of sky and elevation)

2) Expand existing array to 6 panels from 5 panels, (all at same angle and w/ same view of sky) and parallel into existing 250 (two strings of 3 panels), add 6 additional panels at different angle w/ different view of sky and and break into two strings into second classic 200.

Economics not driving key driving metric here (in terms of 5 additional or 7 additional panels, wiring etc either seem acceptable); mostly interested in deriving best efficiency for the system/my time.

I should note that I can reduce shading if need be by cutting tree's, but would prefer to do this on an as-desired basis vs. as a design criteria.

I'm slightly leaning toward option 2 because it seems to be the most robust, gives me the most additional potential KW, and should slightly increase my current 5 panel array efficiency by going to 2 parallel strings at a lower VOC; with that said, expanding my roof top array from 5 panels to 6 panels will involve building more roof (not a huge issue as its a structure purpose built to hold the array and electronics but still a consideration as it could turn this into a two weekend project).

Adding 5 more panels in parallel would be the fastest/easiest but I'm worried that the difference in angle and potential shading could lead to significant inefficiencies. I can't just do 5 into the classic 200 as it would be in HyperVOC all the time.

Thoughts? What is the trade off for adding another 5 panels in parallel  at different angle w/ a potentially different (read:worse) view of sky vs. the over head of running a separate CC on said array?

Thanks for any feedback!

E

[zoneblue]

Personally i would err on the side of less controllers. Heres why:

so long as your arrays are within cooee on vmp, vmp is not heavily influenced by insolation, but more by temperature. As both arrays will be similar temps, the mpp tracking will have no trouble.

More controllers means more idle losses.

BTW your 225 VOC, is that for coldest day of year? And if the array is on  your roof, you will be significantly better off with 150 controllers.

[ebenbayer]

Very helpful to know that vmp is not heavily tied to insolation. That makes this much easier! It sounds like paralleling different angled arrays makes the most sense. Is this true even for major differences? For instance could you have an easterly and westerly array in parallel w/ good results?

What is the resting wattage for a classic? I thought it was in the 10-15W range, but I may have imagined that.

I agree on 150 vs 250; 250 was originally purchased as a back up for hydro controller which requires the voltage headroom.

Thanks for the feedback!

[Resthome]

What is the resting wattage for a classic? I thought it was in the 10-15W range, but I may have imagined that.

More on the order of 3.5 to 5 watts depending on voltage.

See this thread http://midniteforum.com/index.php?topic=153.0

[vtmaps]

How about a third option:
expand to 12 panels, put them on one controller, but use your Classic 200 which has higher capacity than the classic 250.  Put them in four strings, each string being 3 panels in series.  You will need a combiner.

As you know, Classics are more efficient with a lower input voltage.  I generally don't worry too much about a fraction of a percent of efficiency... except when you are running a controller at its limits.  And if you run 3600 watts of panel into a Classic 250 you are pushing the limits. 

Here's a quote from Robin Gudgel (co-founder of Midnite and Outback):
http://midniteforum.com/index.php?topic=324.msg1869#msg1869

Max power for the Classic 250 is somewhere between 3000 and 3500 watts depending on your system configuration.
<snip>
I would not put more than 3000 watts on the Classic 250 myself. It will just run hot. That is never a good idea to run the controller flat out at its maximum day after day.

You didn't mention the distance between your array and the controller... if it is a great distance that might sway you to raise the string voltage, to 3 strings (4 panels in series per string). 

Having your panels in two different orientations is sometimes called "virtual tracking".  There have been a few discussions of that on the NAWS forum.

Try to make sure that the partially shaded panels are all in one string... not one shaded panel in each string.  Also try to make all the panels in each string have the same orientation.

Make sure your combiner has circuit breakers, not fuses.  You can experiment by switching off strings.  You can switch on the partially shaded string and find what it is producing, then switch it off and switch on a string in full sun.  Then you can switch both of them on and see if they are producing less than the sum of their individual productions.  The reason this may happen is that the controller may find a compromise maximum power point.  These experiments will help you know whether it is better to have one or two controllers.

You mentioned that you want your panels to shed snow... one thing that helps with snow is to mount your panels in landscape orientation rather than portrait orientation.  If you mount the panels in portrait orientation, the strings of solar cells in the panel will be in a vertical orientation and when the snow slides to the bottom of the panel it will block a few cells in each string which will reduce your output to near zero.  If you mount the panels in landscape orientation, the sliding snow will block one string of cells, but the other strings will be clear and produce some power.

--vtMaps

[fiddlerkelly]

Zoneblue you stated

BTW your 225 VOC, is that for coldest day of year? And if the array is on  your roof, you will be significantly better off with 150 controllers.

How is the VOC difference calculated from in the midst of summer to depth of winter.

Fiddlerkelly

[zoneblue]

Was just wondering what the temp was when you measured it is all.

Vmp/Voc can go up quite sharply upon cold temperatures. Its the lowest temp you are likely to expect in practice that should be your guide as to whether you will max out your controller.

[inetdog]

Very helpful to know that vmp is not heavily tied to insolation. That makes this much easier! It sounds like paralleling different angled arrays makes the most sense. Is this true even for major differences? For instance could you have an easterly and westerly array in parallel w/ good results?

What is the resting wattage for a classic? I thought it was in the 10-15W range, but I may have imagined that.

I agree on 150 vs 250; 250 was originally purchased as a back up for hydro controller which requires the voltage headroom.

Thanks for the feedback!

There are published studies (by Xantrex and by SMA for Grid Tie Inverters, I believe) that show that as long as neither array has partial shading the purely orientation effect does not cause any MPPT losses as long as the two arrays are otherwise identical (temperature, for example, if the difference is extreme).
The losses will come when there is partial shading, as the activation of the bypass diode(s) in one or more panels will not affect Voc (where the diodes will not be conducting) but will strongly affect Vmp.
Many strongly recommend virtual tracking (E and W arrays) as a way of maximizing the effective sun hours per day, especially in winter in high latitudes, while requiring roughly half the inverter capacity.