diy solar

diy solar

Funny story about the east-west panel spread-out idea

SignatureSolarJames

Try Solar, the Grid will always take you back
Joined
Jun 26, 2020
Messages
1,088
Location
Sulphur Springs, Texas
OK, to get to the point here you will not get some “spread out of power production” by pointing your panels direct east and west. In summary the sum of the gains in the morning for the east panels plus the sum of the losses in the morning from the west panels equal the same production curve total. The same is true for the evening in this configuration. The only thing you affect is the overall output by around 14% for facing off-south with your arrays.

Anyways, I was getting all scientific about this 7 years ago, I split a 10kw array in the PV watts hourly model (available for free!) into 2x 5kw arrays, I plotted out the hourly graph of each 5kw array and thought I was on to something when each array skewed later or earlier in the day.
I then “Summed” the 2 tables into a combined impact table and plotted the graph. Lo and behold it was identical to the standard south facing hourly distribution bell curve :ROFLMAO:

If you know your usage is leaning towards morning you can gain efficiency in direct use by pointing the panels more eastward, if you are heavy in the evening you can cock it west or SW.

Either way the “bell curve” of solar will not be flattened without a tracker chanding the direction in live time. If you overpanel then pointing east and west overall does not help. It may feel like you have “booster rockets” that time of day but franky the net result would be the same but less dramiiic if your “wrong facing” panels at the time were south faced as well.
 
Yeah, the only advantage I can see is if you have a larger air conditioning load in the PM, you may want a slight west bias.
Or if you have scheduled loads, like water heaters or pumps, you may want to have the schedule correspond to maximum panel output.
Or possibly, if you have a PM shade problem, you may want the panels to face the east more, or vice versa for AM shade.
Other than that, it always averages out.
P.S. Just thought of another one.
If you had three arrays, one east bias, one west bias, your solar day would start earlier and end later.
It would average out, but you would reach minimum power-up requirements sooner, and they would end later.
 
Last edited:
I personally would only do it if I was severely over paneling the SCC that I was using. You can put twice as much on one SCC if you position them East/West
 
Without active tracking I would think it might be worth it. Especially if you have a weird roof layout and limited space.

I will be doing normal mount with a westward set to boost it. The main reason for me is simple. Roof space. The southward side will be covered and the westward will carry the remaining panels I'm wanting to add to that section.

Now when I start populating the 2 acre field they will all be south facing for those.
 
the other thing that it may help out is seasonal cloud cover, hazy/foggy mornings become clear afternoons during the summer but can flip in the winter, cool clear mornings then overcast in the afternoon.
 
OK, to get to the point here you will not get some “spread out of power production” by pointing your panels direct east and west. In summary the sum of the gains in the morning for the east panels plus the sum of the losses in the morning from the west panels equal the same production curve total. The same is true for the evening in this configuration. The only thing you affect is the overall output by around 14% for facing off-south with your arrays.

Anyways, I was getting all scientific about this 7 years ago, I split a 10kw array in the PV watts hourly model (available for free!) into 2x 5kw arrays, I plotted out the hourly graph of each 5kw array and thought I was on to something when each array skewed later or earlier in the day.
I then “Summed” the 2 tables into a combined impact table and plotted the graph. Lo and behold it was identical to the standard south facing hourly distribution bell curve :ROFLMAO:

If you know your usage is leaning towards morning you can gain efficiency in direct use by pointing the panels more eastward, if you are heavy in the evening you can cock it west or SW.

Either way the “bell curve” of solar will not be flattened without a tracker chanding the direction in live time. If you overpanel then pointing east and west overall does not help. It may feel like you have “booster rockets” that time of day but franky the net result would be the same but less dramiiic if your “wrong facing” panels at the time were south faced as well.

Need more info. Due East? Due West? What tilt? I've done this, and I get reduced total daily harvest.
 
I expect reduced total daily harvest, because the arrays are aimed optimally for sun taking longer path through the atmosphere.
I expect increased ratio of total to peak. So think it is useful to squeeze more out of the SCC, or grid-export limit, or maximum allowed charge rate.
 
When you are off grid, you often have ample space for panels.

Overall production means very little, getting charge into the batteries as early as possible, and continuing as late as possible will reduce battery requirements. This is the priority.
 
Meh. I call insufficient research and test cases. Varies dramatically by season. 14% is only good as an annual guide, but it's actually closer to 22%.

I just ran it. I tried to pick days near the label that didn't seem to be significantly affected by weather. This is AZ afterall, and sun is awesome.

Here's the Spring equinox:

1684901837167.png

Summer Solstice surprised the shit outta me as the total for the day is a little higher:

1684901702899.png



Autumnal Equinox:

1684901930156.png

Winter Solstice:

1684901537502.png


Annual:

10kW S:
1684902005683.png

5kW E:
1684902040472.png

5kW W:
1684902079104.png

20,221 > (8,344+7,815 = 16,159)

Do better @SignatureSolarJames. You should have run it for more than just summer.
 
It's not instead of south facing.
It's in addition to south facing.
Panels are the cheapest part of solar.


spot on for my system, south roof is full, filling up the west roof, east roof is next, then no more roof. Last will be on top of the pegola on top of the deck on the west side. Small lot, 40x100. Wish I had an acre

Panels are way cheaper now, like a many pencils, relatively. I don't mind "wasting" them so to speak
 
Last edited:
Anyways, I was getting all scientific about this 7 years ago, I split a 10kw array in the PV watts hourly model (available for free!) into 2x 5kw arrays, I plotted out the hourly graph of each 5kw array and thought I was on to something when each array skewed later or earlier in the day.
I then “Summed” the 2 tables into a combined impact table and plotted the graph. Lo and behold it was identical to the standard south facing hourly distribution bell curve

The curve of split E+W arrays will be wider and flatter than the all South array but overall generation will be somewhat less provided there is no restriction on output.

But there are times when a wider and flatter PV output is more beneficial, e.g. in Australia, grid-tied PV systems often have power export limits. You might have 10 kW of PV with an 8 kW inverter but are limited to a max of 5 kW of grid export. If you don't have sufficient load, then the inverter will clip production during the middle of the day.

In those situations it's often helpful to split some of the array towards the East and some to the West. NE + NW here is a really good option. The peak of production for each is about an hour or so apart, and the greater the azimuth split the more the peaks are spread. It also tend to help reduce the level grid voltage rise which can also result in inverter throttling production, or in some cases shut down entirely.

In some locations, e.g. Western Australia, the feed in credit is more than tripled after 3PM. The utility is giving a strong signal that they much prefer solar PV arrays to face West.

Much also depends on when you use power. It is preferably to consume PV directly than to cycle it through a battery (more efficient) and cheaper if it means you don't need as much battery (PV is cheaper). If your household load in the late afternoon is typically much higher than the rest of the day (e.g. Summer aircon consumption) then having a more westerly facing array is going to be helpful.

In many cases though, people are working with the space they have. If you can't build a field array and choose your azimuth(s), then you have the roof you have. Using tilts is typically not that great as it reduces the number of panels which can fit into a given roof area (need to space them out more).
 
And one more:

1684904563559.png

This is a single line plotted for each day indicating S array performance OVER 5kW E + W.

Examples:

Day 79, S outperformed E+W by 35%
Day 162, S underperformed E+W by about 18% (-18% plotted).

Conclusions:

S array vs. E+W array:

22% reduction in annual performance with E+W
E+W can match and even exceed S for a couple of months centered around summer.
E+W performance penalty is worse during all other times.

Most important... RUN YOUR OWN DAMN NUMBERS BEFORE YOU LISTEN TO SOME KNOW-IT-ALL ON THE INTERNET!!!!
 
E-W works well for us, with low tilt. Have too much production in the middle of the day, but its producing useful amount of power from 5am to 8pm at this time of year. Am over-panelled though.
 
Meh. I call insufficient research and test cases. Varies dramatically by season. 14% is only good as an annual guide, but it's actually closer to 22%.

I just ran it. I tried to pick days near the label that didn't seem to be significantly affected by weather. This is AZ afterall, and sun is awesome.

Here's the Spring equinox:

View attachment 150327

Summer Solstice surprised the shit outta me as the total for the day is a little higher:

View attachment 150326



Autumnal Equinox:

View attachment 150328

Winter Solstice:

View attachment 150325


Annual:

10kW S:
View attachment 150329

5kW E:
View attachment 150330

5kW W:
View attachment 150331

20,221 > (8,344+7,815 = 16,159)

Do better @SignatureSolarJames. You should have run it for more than just summer.
Well, I ran it for Texas at 20 degrees, the loss is higher at different and cooler lattitudes definitely. In texas we produce less overall than the radiation would indicate because the heat holds down cell efficiency :ROFLMAO:
 
Last edited:
At $0.55-0.70/W for panels I don't care about maximizing generated energy per installed panel personally. In fact, my main concern is maximizing winter production. Excess summer production is essentially wasted. I am also willing to take the 10% production hit with my panels back-sloped to the north 5°.

But, that is where design comes into play.
 
Well, I ran it for Texas at 20 degrees, the loss is higher at different and cooler lattitudes definitely. In texas we produce less overall than the radiation would indicate because the heat holds down cell efficiency :ROFLMAO:

Nope. Wrong again. I'm at slightly higher latitude and about 15°F cooler. PVWatts factors cell temperature in. Why do you think it reports Wind Speed and Cell Temperature?

I ran it for your address @ 20 degrees. You're only about 1.3° more South than what I ran above.

10kW S:
1684936269178.png

5kW E:
1684936301985.png

5kW W:
1684936329324.png

12,913kWh split vs. 17,213kWh South. That's a 25% loss annually - HIGHER than what I reported above, AND higher than what you claimed at 14% - likely due to 20° vs. 34.4° tilt.
 
Last edited:
Back
Top