Looking to go solar...everything looking good, am I missing something?

[HX_Guy]

Also just eyeballing your aerial photo, if that is indeed on Nov 6 mid-morning, you can just see the neighbor's roof shadow start to clip the bottom of your roof, which means at noon it's not shading at all. So you definiltely have no shading issue at all before Nov 6 and conversely after Feb 4 or so, that's your worst case. More likely it's much better than that, and maybe there's no shading at all on the higher of the two rows.

FWIW, I followed the shading of my arrays by the neighbor's eucalyptus trees (about 100 feet away, but they're like 70' tall!). The shadows started to obscure around Nov 17 for me, by Jan 25 I'm all clear. For the upper of my two rows, the shading is only from Dec 1 to Jan 10 or so, nearly a month less than the lower row. During those days I still get no shading from those trees until mid-day, so I can still get a half-day of production.

What might annoy you more is that tall vent stack (pipe) on top of your neighbor's roof peak. That looks like shadow would sweep over a couple of panels during most of the day, and for several more months than the roof. If that small shadow covers just part of a panel, it can negate the production for the entire panel.

It turns out that Nov 6th is not the date of that photo, it says 11/6/2012 on the website but it must be referring to something else because our house was a dirt lot on that date. I do know that it was Sept 2013 - Nov 2013 though as it says so from the drop down menu (and I can choose other years too, but none help me any). If my Sketchup model is any good, it would appear the aerial is from early September as that's when the clipping is about the same during that time of day.

Problem with my Sketcup model though is that it's not exactly to scale. It's close, but my roof pitch is not accurate and neither is the width, it's off by a 2-3 feet which can make a big difference I assume. That's why the more I think about it, the more it's probably best to wait and take some photos this upcoming winter. I wish I would have done it last winter but it didn't even cross my mind that it would be a problem.

[wwu123]

I do not think that the app shows you a picture of your neighbor's house.
Instead you use the camera on the phone to put up a picture looking from your panel location toward your neighbors house and look to see where the sun trajectory imposed on that picture by the app crosses your neighbor's house, trees, etc.

Incorrect, the app shows you the neighbor's house in realtime using camera view on the phone - assuming you point the iPhone camera at the house. And it superimposes the sun path automatically. Ah heck, let me show you...did I forget to mention it will take a snapshot for you?

This is a photo taken in February around 9 am, showing the path of the sun on March 20, plus where the sun will be at 3:40 pm. The sun is not actually in the photo, it doesn't work well to have the sun in the viewfinder - the bright disk is superimposed for 3:40 pm. When you manipulate the app in realtime, if you line up the sun disk with the sun path and the edge of the obstruction, you will find out when the sun gets blocked and by exactly what. You can project the path for any day of the year no matter what day it is today.

South_0320_1540_farside2.jpeg

In this case I think it was around 3:50-4 pm for that tree. No that's not a eucalyptus, although you can see one of the aforementioned eucalyptus tress towards the lower left. Yes, I do have multiple tree shading issues, sadly, on both sets of array.... hence why I've spent a lot of time with this particular app standing on my roof.

EDIT: inetdog, I think you might've been saying the same thing actually, but I initially interpreted a bit diferently....

[HX_Guy]

Pretty neat...so basically if you see the sun still from where you set your iPhone, then you're ok, if not, then it's in the shadow of whatever it's pointed at, correct?

I guess I thought it was more of a you view the object you want to see if it's shade or not...so I would be on the edge of my roof, pointing the phone toward the roof and in the viewfinder I would see the shadow path. That would of course only work if it was using both the front and rear cameras to see what's in front and behind the phone at the same time. Any good programmers out there?

[JCP]

Losing 50% (or whatever the real number is) of your solar production during November-Feb may not be that big a deal since those are the months with the least amount of sun. Looking at the insolation numbers in Phoenix, November through February represent about 21.5% of the whole year insolation. So losing 50% for those 4 months would impact the total annual production by only slightly more than 10%. I know that this is very ballpark and requires more data, but that may help put things into perspective.

[wwu123]

Pretty neat...so basically if you see the sun still from where you set your iPhone, then you're ok, if not, then it's in the shadow of whatever it's pointed at, correct?

I guess I thought it was more of a you view the object you want to see if it's shade or not...so I would be on the edge of my roof, pointing the phone toward the roof and in the viewfinder I would see the shadow path. That would of course only work if it was using both the front and rear cameras to see what's in front and behind the phone at the same time. Any good programmers out there?

Close but not quite. You want the phone to be physically where a panel will be - ideally get a ladder and get on the roof, and squat on the roof where a panel might be, facing the neighbor's house. If you can't get on the roof, get on a stepladder next to the house, putting your head and the iPhone about the height of where your raingutter would be, basically about where the bottom of the array would be. You want to use the rear camera to get the potential obstruction in view, and it will draw out the sun path, so you can see if the sun crosses behind the obstruction and when it does.

You will not be pointing any camera at your own roof where the panels will be, because you should be standing on or near that spot. It's really quite self-explanatory once you do it, and you need no other calculations or software. Because the potential obstruction is so close, the angles change a lot though every few feet. So the only hassle is, you'll need to repeat this every few feet where a panel might be, because the angles are very acute and will change dramaticaly - but that's because the shadows will change dramatically.

Also you don't want to do this standing straight up on the roof, as you'll be six feet above the panel locations, again changing the angles dramatically.

[HX_Guy]

Ok coolest app ever! Damn pretty impressive technology for a "phone".

Of course I got up on my roof and did some testing really quick. I took the photo from as low down as I could while still being able to see the screen and tap the photo button. One thing I'm not sure of though is the angle and how does that affect things? I mean sure I can see the sky, but wouldn't the building's shadow get longer as the sun is lower? These photos were taken with the phone basically at a 45º angle while the roof is at a 14º angle, if I took the picture at 14º, all I would see if sky of course.

Here are the pictures, I set it to December 21st at 12:30pm.

First photo is from 3' down from the roof edge which is where a panel could start...


Second picture is from 5' lower, figuring a 3x5 panel (and also 5' from the roof bottom, roof width is 13').



Based on these, what's your guys' take? And again, what kind of affect does the angle that the phone was at while taking the picture play into it?

[wwu123]

While you're on the roof, if you center the realtime image over the bright "Sun" disk (raise the phone slightly until that faint circle outline below the Sun image is centered on it), it wlll suddenly show you the entire path of the Sun during that day. And if you then move that faint circle along the path, you'll see where the Sun will be not just at 12:30, but at different times.

What you'll see is that after 12:30, the path goes to the right and down, until at some point it crosses from the clear sky to below the neighbor roof. Let's say for the first photo that happens at 1 pm - after that, from where you're standing you can't see the sun, and the sun can't "see you". The spot you're standing on will be shaded on that day - that is exactly the roof shadow getting longer as it gets lower in the sky and reaching your location.

For the 2nd photo, taken lower on the roof, if the Sun is just peeking out at 12:30, then what you'll see is that most of the arc of the sun is below the roof all day, only peeking out at high noon as projected in the photo. So basically you'll get no production all day on panels that low, at least on the Winter Solstice.

Now if you repeat both experiments standing about 20' to your left, where you'd be looking straight out at the left edge of your neighbor's roof, you'll probably find that both the higher and lower panel locations can see the sun fine in morning because the sun is still to the left (southeast) of the neighbor's house. So you'll get a half-day or more. Similarly if you move to the far right of your roof, you'll get sun in the afternoon or more. So you could split the arrays to the far left and right of where you were standing, avoiding the "donut hole" in the middle bottom that sees the worst of the winter shadows.

[fuzzelogic]

I would recommend staying away from the fronius inverters, people I know who have used those inverters have pretty much all had to have them replaced multiple times over the last couple of years.

Yeah, the fronius support will let you send them back without much of a issue, but the inverter will always seem to die when you least want it to go... In the summer during peak production when your getting ready to go on vacation.

SMA sunnyboy is another inverter and those seem to be built like tanks as my friends are still on their original inverters some of which were installed at the same time as the Fronius.

The systems with the fronius inverters have been installed with various panels and locations, so it's not all isolated to a single location where these inverter downtime issues have occured.

I'm not familiar with other brands, but I would have a hard time recommending those to you as a forum member friend. As a installer, I'm not sure why they would want the headache of having to always service them??

Take what I say with a grain of salt, but if you must have fronius, I would check with some your installers previous customers to see if they are "happy" with those inverters!!!

Also, with that much shading from the neighbors roof would be horrible for string inverter output and you would probably need micro inverters if that shading is as bad from what your mockups show!!!!

[HX_Guy]

Now if you repeat both experiments standing about 20' to your left, where you'd be looking straight out at the left edge of your neighbor's roof, you'll probably find that both the higher and lower panel locations can see the sun fine in morning because the sun is still to the left (southeast) of the neighbor's house. So you'll get a half-day or more. Similarly if you move to the far right of your roof, you'll get sun in the afternoon or more. So you could split the arrays to the far left and right of where you were standing, avoiding the "donut hole" in the middle bottom that sees the worst of the winter shadows.

What you're describing here sounds like a plan that one company suggested, thought they still had panels in the middle, but they had more on the ends around the "donut hole".



About string inverters...is some of the panels are shades, does that mean all panels on that string are not producing anything? So even if I had panels just on the left and right, and the left side panels were shades in the morning, would the right side panels still be good for anything? Or basically both sides would only be good during mid day when they are both in full sun?

[silversaver]

If one panel is affected, the whole strings act the same.

[HX_Guy]

If one panel is affected, the whole strings act the same.

Well that sucks. I guess micro-inverters would be the better way to go?

[silversaver]

Well that sucks. I guess micro-inverters would be the better way to go?

You can use the inverter that has dual MPP trackers on the shadow side of panels. Take a look of SMA TL-US series that has OptiTrac (shade management)

http://www.sma-america.com/en_US/products/grid-tied-inverters/sunny-boy/sunny-boy-3000tl-us-4000tl-us-5000tl-us.html

[HX_Guy]

One of the companies mentioned the dual MPP thing but they were wanting to put all south facing panels together and then the east facing together (something about 14+14 south facing panels and the remaining east ones on a string). That doesn't seem like it would be ideal to me as then the south facing panels on the north end of the house would be affected if the south side panels are in the shade. Maybe they are putting more importance the rest of the year and not the 3-4 months during winter.

[inetdog]

You definitely do not want to put panels with different facing directions into the same string. It will really reduce your output.

And if you can get the same number of panels (all one direction) into each string, say 1 E string and one S string, then IF neither string has partial shading issues you can combine both strings into one MPPT input with almost no losses at all.

But if the shade is going to be marching across the array during any hour of decent production, you will be better off with either micro inverters or DC optimizers.

[fuzzelogic]

Another possibility would be to use 2 inverters. 1 for the large roof and 1 for the other smaller system.

1 - it gives you be better redundancy if 1 of the inverters would have a issue.

2 - It could give more flexibility in not having all the panels tied to 1 inverter.

The catch would be is if you would have enough panels to turn on the smaller inverter...

When looking back at earlier posts, it looked like both quotes only mentioned 1 inverter?