Help with orientation??

icarus

Reading through another site, I came across the following. (I'll post a link at the bottom of this post) The owner was proud of his new $40k Pv system and had posted pictures on another forum. I was instantly struck by what he had. I would like the opinion of those who live in more southern climes with their grid tie systems.

Picture this, A tiled roof house, with a 3/12 (~14degree) pitched roof). The ridge line runs ~ 40 degrees East of due south. The location is 33.38 N 117W.
One this roof are 36 225 watt panels for a total of ~8kw. These panels are place on BOTH sides of the ridgeline, with no tilt for elevation. My question for the number crunchers out there is, how much efficiency is this system going to lose as a result of the panels not being ideally alined?

My concern (with only limited knowledge of the reality of the situation) is that this guy spent ~$40k AFTER rebates, tax credits, and incentives for this system, and it seems to be a not very efficient use of his resources.

I would be curious to see what you guys think,

Tony

http://priuschat.com/forums/environmental-discussion/55486-cleaning-solar-panels-4.html

Posts #33 on are the thread in question. There are pictures of the array, as well as a satellite picture of the house. I couldn't just cut and past the pictures, hence the link.

Cariboocoot

Re: Help with orientation??

I'm inclined to agree with you - it's not very efficient. (Sorry about the pun: couldn't resist.)
Way up North having panels on the 'back side' could work in summer: at 51° lat the sun rises on the back side of my cabin in summer. (I've also considered putting panels on the porch roof to catch morning light, as that gets lit long before the sun clears mountains/trees to the main roof.) The closer to the equator, the less difference between summer/winter angles.
Roughly speaking, the ones on the 'front' are 10° 'too much' for summer, and about half what they should be for winter - maybe a good year-'round compromise.
But the ones on the back side would be angled "-14° from 0°", as it were, so it looks like they'd run about 40% efficiency?
This was obviously an installation compromised made necessary by the site limitations. Probably by-laws or neighbors would object to mounts giving 'true south' orientation. I hope he has a separate controller on each bank, and has a look at how much each puts out on a good, sunny day.
Compromising efficiency for aesthetics is throwing away money.

icarus

Re: Help with orientation??

In this case, it is not just the 'backside' panels that are not optimum, but the panels on neither side are optimum, the backside panels are even worse. Both sides are off plane in two dimensions, (by a lot it seems) but the back ones are worse. I routinely move a couple of my panels to take advantage of real early morning sun in the summer. In the winter I get shaded quite early, so I move two of the panels out onto the ice to get the afternoon sun. If I am feeling like it, I reorient them throughout the day, but my array is small enough to do that to. (2 60 watt panels that I can move, 2 that are fixed)

My real concern here is that the customer relied on the "expertise" of the installer/sales guy. IMHO, unless it was with a clear "buyer beware" that the panels would not perform as expected, then I think he may have gotten a bum deal. I would like to see someone's (Bill, SG, Jim, etc) calcs and opinions on the subject. It just seems that if you spend upwards of $40k you darn well should get it with your eyes open, and a maximum yield.

Tony

solartek

Re: Help with orientation??

I don't think the situation is as dire as everyone is making it out to be. There is enough information provided to perform some sanity checks using an online estimator.

Let's run some calculations using a tool everyone can use: PVWatts. Based on his location it looks like Long Beach, CA is probably the closest location to use in PVWatts. He has about 4kW on each side of the roof so let's work with that size array to generate some annual kWHr predictions for each side of his roof. First do a baseline calculation. Assume a 4kW array facing due South (180 deg) and tilted at latitude for Long Beach (33.82 deg). PVWatts predicts an annual production of 5795 kWHrs.

Now decrease the array tilt down to his assumed roof angle (2/12 or 16.67 degrees). Also orient the array so it faces SouthWest (225 degrees). What does PVWatts predict? An annual production of 5597 kWHrs or about a 3.42% decrease in annual production over the optimal situation of due South and tilted at latitude. The flatness of his roof helps dramatically with the misorientation from due South.

The more distressing decrease is when the array is facing NorthEast (45 degrees). Here PVWatts predicts an annual production of 4334 kWHrs or a 25.2% decrease. Bad, but not catastrophic.

The real issue with this system is if there is only one inverter being used. There is no way he is going good performance of the MPPT in the inverter with two arrays so mismatched in orientation. Treating the panels on each side of the roof as a separate subarray with it's own inverter will maximize his possible energy harvest.

Edit: I went back and looked at his pictures and he does have two inverters for this system. Assuming each inverter handles panels for one side of the roof, then he has the best possible situation for his setup. You could always add more complicated tilt-mounts but you need to weigh the added expense versus the predicted increase in annual energy production.

Scott.

icarus

Re: Help with orientation??

Scott,

Very nice. I do have a couple of questions however.

"Now decrease the array tilt down to his assumed roof angle (2/12 or 16.67 degrees). Also orient the array so it faces SouthWest (225 degrees). What does PVWatts predict? An annual production of 5597 kWHrs or about a 3.42% decrease in annual production over the optimal situation of due South and tilted at latitude. The flatness of his roof helps dramatically with the misorientation from due South."


The way I read this is if you assume directly south, tilted 33.82 degrees for lat, gives you your base line of 5795 kwh/yr. The problem I see is that you are not dropping the angle to only 14 degrees, you are really, in effect laying them flat, and then raising one side 14 degrees, so that the panel is way off plane relative to "normal".

Does your analysis take into account that the panels are on a ~14 degree pitch side ways to south. In other words, they are 40 degrees off south in one plane, and rather than being tilted up for latitude on that plane, they are pitched off plane 14 degrees. (I realize reading this I am not asking the question properly, but I hope you know what I mean.

If your analysis is correct, you would predict about a 30% net/net loss over the course of the year? (the 25% plus the 3%) As you say, not very good, but not real great but not as bad as I imagined.

Being a far northerner, I sometimes have a hard time fathoming life so far south.

Tony

niel

Re: Help with orientation??

scott,
you might want to refigure that as 1> the pitch was listed as 3/12 and not 2/12 and 2> the angles for a roof facing east/west would be 90 degrees to the east and 270 degrees to the west. at best this will perform as a pv system of half its size during the summer and during the winter it will be even less because neither bank is towards the south enough. on edit i see it is offset by 40 degrees and this may help the winter production for 1/2 of the pvs, but because the pvs face 180 degrees from one another the general statement i made is still true that this will perform at best to be 1/2 the system size as it is.
this would also be better organized in my thoughts, but i'm having connectivity issues with my dsl signal fading in and out.

caribootcoot,
the closer to the equator (at equinox) you go the larger the height angle variation with the sun going nearly straight up overhead from the east and back down to the west as it is in the north (summer solstice) that it will stay close to the horizon, but with nearly a full 360 degrees around. everybody else in the northern hemisphere is somewhere between the 2.
to sum it:
at 90 degrees north lat the summer solstice the pvs will be aimed at the horizon, but would need to turn a full 360 degrees to continually capture the sun.
at equatorial equinox, or 0 lat, the sun rises in the east (90 degrees) and goes straight up to 90 degrees in overhead elevation and straight back down to set in the west (270 degrees).
all others are a combination of the 2 to various degrees and depends on latitude and the time of year making a unique solar curvature for every latitude on the earth.

icarus

Re: Help with orientation??

Now I am really confused!

I ran the numbers, and I don't think Scott erred . 4kwdc should put out 5795 kwh/yr if properly oriented. The east facet array (180-40-90)facing 50 degrees and tilted 14 degrees would put out ~4508kwh. The west facet, facing 230 degrees would put out 5518 kwh. The total combined, under ideal orientation would out out ~11,560. These two facets together will yield, 10,026 kwh/year or about 86% of perfect.

There seems to be something amiss in these numbers. My intuition says the yield should be lower. Perhaps it is my northern bias!

Tony

Solar Guppy

Re: Help with orientation??

I took a look at the pictures... The bigger issue is shadows, the roof has walls within a foot or two that cast shadows for a large part of the day, if you scroll down in the link, there is a picture from above that clearly shows the shadows

There is also stacks and things ... clearly not an ideal roof for PV

icarus

Re: Help with orientation??

Just to illustrate my northern bias, I ran the numbers for International Falls MN. It shows on a similar roof near me, the losses would be ~30%. The winter harvest would be terrible, but the summer would be quite good since it the sun is so high over head for so long.

Personally I have had to optimize my system for winter days, but interestingly enough, our production is quite good because of snow reflection and cold pv temps. Our summer production is hampered by trees, but our loads are much less since it stays light until 11 PM!

Tony

solartek

Re: Help with orientation??

Several people to answer.....

icarus
With a tool like PVWatts (or any other simple solar estimator) your two main knobs are what angle you are going to tilt the panels up from flat (Array tilt) and in what direction are you going to tilt them (Array orientation). For PVWatts inputs, flat panels are tilted at 0 degrees and panels oriented due South are at 180 degrees (i.e. normal compass headings). The baseline example I calculated assumed that the panels were tilted up 33.82 degrees from flat (latitude at Long Beach, CA) and that the panels were tilted up in the due South direction (180 degrees). This is the optimal position we all know so well. Where things get confusing is when we deviate away from that position. For my examples, you could think of the ridge of the roof running along a NW-SE line. One face of that roof would face SW (225 degrees) and the other face of that roof would face NE (45 degrees). These are the array orientations. The tilt of the array came from assumed pitch of the roof being 2 over 12. The original poster actually thought his roof was less than 2 over 12 which would be even flatter. I think he also stated his roof ridge ran only about 40 degrees off of N-S where I assumed 45 degrees. Neither difference creates a different conclusion from my original one.

niel
As I stated above, the way I read the original thread the poster thought his roof was actually flatter than the 2 over 12 I assumed. As for your contention that an array facing due East or due West would perform only at 1/2 it's size during summer and worse during the winter, I respectfully disagree. It's way more complicated than that. For arrays oriented off of due South, the tilt of the array has a major impact on the overall production. The less the tilt of the array, the less impact that orientation has on the overall production. The limiting case is, of course, for flat arrays (tilt = 0 degress) where the orientation doesn't make any difference. For my examples a 16.67 degree tilt is fairly small, so for the SW facing array which is only 45 degrees off of due South, the impact caculated by PVWatts was minimal (3.42% decrease).

Solar Guppy,
I agree completely with your concerns about shading from the wall adjacent to the southwest orientated roof. Hopefully a shading analysis was done and the impact of it was taken into account by the designer.

icarus (again)
There is nothing wrong with your calculations. The interaction between array tilt and orientation is not typically explored enough. Particularly for shallow roof pitches (and therefore array tilts for roof mounted installations), the decrease in annual production of an array not facing due South is usually overstated.

Scott.

icarus

Re: Help with orientation??

Scott,

I think you got it covered.

I spent some time playing with PVwatts, using different angles with different latitudes. It is interesting how it changes as you go north (or south) from the equator. I have lived my life in the north, short winter days, low sun angles, long summer days with high sun angles. Looking further south is a little bit eye opening. My experience has been exclusively with off grid PV and flat plat hot water.

Thanks one and all for your help.

Tony

PS I too saw the shading,,,