PV panel orientation question

khotton · January 2009

Hi experts,

If one's house faces south but the ridge line runs South-to-North with a 4/12 pitch with a grid-tie system is it possible to have two arrays (say 1.5kW each), one on the east sloping roof face and the second on the west sloping roof face and still efficiently use only one grid-tie inverter?

Kevin

mike95490 · January 2009

Re: PV panel orientation question

I * think * it would work, as each array's voltage should be the same, except maybe at dawn and sunset. The lower producing array should not affect the inverters MPPT tracking.

icarus · January 2009

Re: PV panel orientation question

Go to PV watts calcualtor :http://rredc.nrel.gov/solar/codes_algs/PVWATTS/

And do two calcs. One for the east facing roof facet, and one for the west facing. I think (Intuition) you will find you get a much bigger yield with two inverters.

You should also look at your site in the real world. Do you have early morning fog such that you get sun later than the calcs would show? Or do you have something that favors the afternoon side? If so, you could consider putting more panels on the favored side and fewer on the least favored side.

Good luck,

Tony

BB. · January 2009

Re: PV panel orientation question

Yes, as long as one series string is on one roof surface facing east, and the second string is on the other facing west. And, at "noon" the total power from both strings is less than the maximum rating of the GT Inverter.

Fronius USA has a nice white paper that descusses the impact of East/West strings on one vs two GT Inverters:

White Paper Home page

FRONIUS IG - Reaction to Non-Optimal Conditions (PDF download)

Short answer is that East/West strings on one inverter vs two inverters, less than 1% difference between the two.

So, just look at your costs (1 big vs 2 small inverters) and purchase which ever is cheaper and easier to install.

-Bill

icarus · January 2009

Re: PV panel orientation question

I'd do a PV watt calc: http://rredc.nrel.gov/solar/codes_algs/PVWATTS/ one using the east facing array, and one with the west and see what the numbers reveal.

I would also look at the local, real world site condition. Does the east side get sun late due to common morning fog for instance? Does the west facet get shaded due to afternoon fair weather cumulus clouds? You could then consider favoring one side or the other with additional or fewer panels.

Tony

BB. · January 2009

Re: PV panel orientation question

Using the Link that Tony/Icarus supplied and trying Boulder Colorado:

3kW of solar panels (defaults, fixed mount) = 4,376 kWhrs/year
1.5kW south = 2,188 kWhrs
1.5kW east = 1,723 kWhrs
1.5kW west = 1,520 kWhrs
1.5 E+W = 3,243 kWhrs
1.5 E+W / 3 South = 74% of south facing array

All things being equal, fixed array, accounting for weather, and for temperature. One or two inverters, should not make more than 1% difference.

-Bill

icarus · January 2009

Re: PV panel orientation question

Using Boulder,

An 18 degree roof pitch (4/12), east orientation of 270, I get annual yield from 1.5 kw of 1702 kwh/yr

Using 090, 18 degrees yields 1816 kwh/yr

A combined total of ~ 3518 kwh/yr

A ideal set up should yield (3 kw, 40 degrees tilt, 180 south orientation, 4376 kwh/yr

So you can see that you are running ~80% with such an installation.

Tony

khotton · January 2009

Re: PV panel orientation question

Thank you for such prompt and thorough replies. I will go and study the white paper on single vs dual inverters.

Kevin

BB. · January 2009

Re: PV panel orientation question

Interesting results from Tony plugging in your real slope... The East/West panels mounted at 18 degrees generated a bit more than when mounted at the 40 degrees that would normally be done for a south facing panel array.

It would appear that playing with the solar calculator is worth the time to ensure that optimum panel placement is obtained.

-Bill

Moe · January 2009

Re: PV panel orientation question

BB. wrote: »

Interesting results from Tony plugging in your real slope... The East/West panels mounted at 18 degrees generated a bit more than when mounted at the 40 degrees that would normally be done for a south facing panel array.

It would appear that playing with the solar calculator is worth the time to ensure that optimum panel placement is obtained.

-Bill

Raising the angle of panels facing due east and due west to 40º tilt in those directions will significantly reduce the time the panels are getting sunlight through a thinner atmospheric path... and should result in a loss as you discovered.

Assuming the OP is at 40º latitude, the north sides of the panels on both the east and west sides of the roof should be elevated, so the panel tilt is 40º southerly, including the 10/12 pitch in the east or west direction. Simple solar calculators aren't going to give you this info no matter how long you play with them because they can't deal with tilt in two planes.

BB. · January 2009

Re: PV panel orientation question

This calculator does the calculations correctly--it is just that the normal suggestion of placing the panels at latitude angle for optimum power is not correct if the panels are not facing due south.

It will just take some trial and error to find the optimum angle for arrays not facing south.

I would also guess that some times, even panels the face due south may not be optimum "power output" at latitude even when facing south when other issues are taken into account.

For example, during the "summer", my Time of Use Net Metering rate plan pays me ~$0.30 per kWhr weekdays between noon and 6 pm.

Otherwise, the other six months of the year, they pay me $0.11 per kWhr for noon-6pm and all other times (winter/summer) $0.09 per kWhr.

I would assume that if I place the panels at some angle less than latitude, that I would find an "optimum" angle which would generate me more money based on my summer billing--even though it produces less power over the 1 year period.

The interaction between panel output over the season, rate plans, and, possibly even what time I use power during the day/season--is all going to affect the optimum panel orientation vs $$$$ output.

And, then in the next few years, there is a good chance we will be forced to go to Real Time Billing... Then, we will not even be able to predict when our "optimum" generation periods are--because the pricing will be purely based on "supply and demand" with probably a 24 hour forecast.

In the end, there is no possible way of accurately predicting solar panel system output in $$$$ vs return on investment because the is no predicting a major factor factor in the cost equation--the cost of power during the ~6-8 hours of the day I am producing power.

Anyone interested in investing $30,000 or so (before rebates) in solar / renewable energy in this current market?

GT solar will still help on the bill--but the unknowns just make the whole decision process even more difficult.

-Bill

icarus · January 2009

Re: PV panel orientation question

It is interesting, and important to look at the yield over the months. Clearly an off angle roof (fixed) array is going to have limitations. If you can get better than average harvest in the summer at the expense of the winter, for example, and you are net annual metering, then that might be the way to install. If, on the other hand, you have a pressing need for a good average 12 months, (off grid for example) you might choose to array differently.

I too have a hard time understanding how PVwatts takes into account tilt in two planes. I assume that it does. On another thread a few weeks ago we had a similar discussion from a use in Santa Barbara. Playing with the tilt angles, from 0 (flat) to the proper seasonal angle confirmed in my mind that the program does factor in two plane tilt.

Having said that,, in the example in this thread, the fixed tilt, dual array, east/west, projects a harvest of ~80% against an ideal tilt angle/azimuth fix tilt.

With 1 axis tracking yields ~5707 kwh/yr
With 2 axis tracking, comes in at at total annual harvest of ~6065kwh/yr

I think that it is both instructive and informative to look at the monthly harvests, and play around with optimum angles and see what the monthly results would be. I also reiterate the need to take into account local weather averages and quirks in siting and tilting.

In my case, I use a vertical tilt (flat against the wall) for the dead of winter. The advantage I gain is, the panels get now snow coverage, AND, get better reflection off the considerable snow cover. This is an example of letting common site sense have a role over strict numbers.

Tony

Moe · January 2009

Re: PV panel orientation question

I assume that even with a North/South ridge, one would want a south-facing tilt... in which case there would be a compound angle affecting the tilt:

icarus · January 2009

Re: PV panel orientation question

A problem with tilting panels on off facing facets (is that bad wording or what!) can be shading issues. For example, it you have a east facing roof facet, and you tilt the panels to some optimum tilt angle they then (could) cast a shade on the panel behind, especially in higher latitudes in the winter. If you are short of room it could be a problem because you might not be able to space the panels far enough apart to keep on from shading another.

T

Cariboocoot · January 2009

Re: PV panel orientation question

This brings up an interesting point:
You can't really know how your panel site will work until you've looked at it all day, every day, for a year. It's just amazing how the site that looks perfect on July 21 at noon turns out to have disrupting shadows 80% of the rest of the year. The further North the worse it is as the sun angle changes drastically over the course of the year.

In this case I think I'd put the panels as near to the ridge as possible and use separate controllers for each side. A bit extreme and probably not the best choice but ... It's hard to tell someone anything about their site without seeing it.

BB. · January 2009

Re: PV panel orientation question

The help page for the PV Calculator discusses many of these issues (and appears to factor in just about anything that matters).

They also have a graph of "packing density" of solar panels vs tilt.

And since the program is based on actual data for ~20 years or so--it should take into account the "average weather" at the site...

However, if your location has significant weather difference vs the reference site--a bit of local tweaking may help.

-Bill

icarus · January 2009

Re: PV panel orientation question

Cariboocoot wrote: »

This brings up an interesting point:
You can't really know how your panel site will work until you've looked at it all day, every day, for a year. It's just amazing how the site that looks perfect on July 21 at noon turns out to have disrupting shadows 80% of the rest of the year. The further North the worse it is as the sun angle changes drastically over the course of the year.

In this case I think I'd put the panels as near to the ridge as possible and use separate controllers for each side. A bit extreme and probably not the best choice but ... It's hard to tell someone anything about their site without seeing it.

On the other hand, if you are in the planning stages, some judicious use of compass and inclinometer, p-level, or builders level, armed with you latitude info, you can site various elevations and azimuths and get a pretty good idea of where and when the sun is going to fall, and where and when it won't. I was surprised when I build a new house (cabin?) on he same island I have been on for more than 50 years. It turned out, contrary to my own perception, that it loses sun much earlier in the day in the summer than I imagined, but keep sun much longer in the day in the winter. It also gets 100% sun from sun up year round. I adjust my panels according to season (sometimes!) Since I have greater sun in the early morning, AND in the summer have consistent afternoon fair weather cumulus clouds,,my morning bias actually works to my advantage,, counter to initial calculations would have suggested.

Tony

khotton · January 2009

Re: PV panel orientation question

BB. wrote: »

The help page for the PV Calculator discusses many of these issues (and appears to factor in just about anything that matters).

They also have a graph of "packing density" of solar panels vs tilt.

And since the program is based on actual data for ~20 years or so--it should take into account the "average weather" at the site...

However, if your location has significant weather difference vs the reference site--a bit of local tweaking may help.

-Bill

In reviewing the help page, I was struck by the figure showing 1-axis tracking. What is shown would actually be equivalent to 2-axis tracking (it is shown rotation about a inclined axis -- like an equatorial telescope mount). I thought 1-axis trackers tilt the panels at the site's latitude and rotate in azimuth (as such 1-axis trackers because less effective near the equator). I'm I correct in my understanding?

Kevin

BB. · January 2009

Re: PV panel orientation question

I guess there are many types of trackers out there... This link probably lists all of the basics as you know them.

The single axis polar mount tracker would tend towards a horizontal mount at the equator and a vertical mount in the high latitudes. I would guess, depending on the intended destination (and where they are designed and built), different companies might choose to solve the the tracking problems in different ways.

The polar mount (angled to latitude) would be ideal, but it still does not track seasonal differences (the earths axis is relatively fixed to the sky, but not in relation to our orbit around the sun) and a two axis tracker will give better results (or adjusting the polar mount elevation a couple times a year).

The WattSun seems to be an azimuth drive (for 1 axis panels) as you surmise--but, I would guess that it is best then for the more polar latitudes... And the Solar Calculator would not accurately predict its performance near the equator (and in fact, would just "roate" the flat panel array for no help at all over a fixed array pointing straight up).

But then, WattSun also appears to offer horizontal axis trackers for large commercial options.

Presumably each has their advantages and disadvantages.

On this site, we normally suggest that people compare the output and cost of installation (and maintenance) of a tracker array with a fixed array (possibly with tilt--especially useful in snowy areas to tilt panels near vertical for better snow shedding).

In the end, with the relatively low price of solar panels today, it is either cheaper or a wash just buying more panels and a fixed mount that to use a tracker.

Do you have specific requirements that need a tracker?

There are people here with much more experience than I (who has virtually none) with trackers.

-Bill

PV panel orientation question

Comments

Categories