Solar Panels, Leafy Shadows and Bypass Diodes

[CA_Traveler]

3x250W panels, 3 bypass diodes each, series connected 91Vmp. A 5-10A battery draw over the panel output was maintained until 2:08. The rig is in Casa Grande, AZ about 40 miles south of Phoenix and facing south on 3/5/15.

My interpretation of the graph and pictures. The tree shadows on the front 2 panels up to 12:48 caused their bypass diodes to conduct. The rear panel had full sun (see 7:30 picture) until 8:08 when there was a shadow on the leading edge (see 9:38 picture) causing one of it’s bypass diodes to conduct. At 10:08 the shadow was gone. The rising sun caused the rear panel to increase it’s amps and it peaked at 15A (see 10:30 picture).

At 12:58 the leaf shadows start to leave the front 2 panels and the effect of their bypass diodes is very evident. The voltage peak at 12:58 may have been due to wind. The voltage shows sun returning to one panel as the voltage reaches 60V as the bypass diodes are no longer needed and the amps continue to ramp up. This continues with the third panel up to 1:38. At 1:38 both the voltage and amps are at full output. At 1:38 and 85% SOC most of the house loads were removed and normal battery charging continued. The 2:18 transistion was likely normal house usage and the declining amps clearly show the battery SOC approaching full charge. The 2:18 voltage reduction is due to the MPPT reducing the charging amps.

At 12:58 the sun elevation was just pass it’s peak and changing slowly which combined with the 15 degree/hour azimuth change cleared the shadows from the front panels in about 30 minutes.

What’s your take on this data?

[2oldman]

Not all shadows are the same, (as Salvo points out) and not all bypass diodes work the same.

I have a BP panel that, when shaded by my A/c cowling even slightly, will degrade my array's performance drastically. Now, if I cover the panel completely, it's as if it's not there, which is perfectly fine. I had to fashion some alternate support brackets to get it out of the shade. Its tilt angle is different from the other 5 panels, but that's of no concern.

[CA_Traveler]

MSView is a Morningstar program that creates a csv file. The program has a graphing option but I used Excel. With Excel I converted the UTC time to a more human oriented time and added the right hand axis which then gives two different vertical scales.

You choose the variables and sample time for the program and the above graph used a 30 second sample time.

[2oldman]

Good writeup. How was the graph created?

[brulaz]

You were smart to spread your panels fully over the roof so hopefully at least one could avoid shadows.

My two parallel panels are both at the rear of the trailer. And often when I back into a site, I back into shadows covering both, at least in part.

Discovered last summer that I got more solar energy at an open site on an overcast day, than at a partially shady site on a sunny day.

My next project is to add a third parallel panel at the front of the trailer.

[1492]

Moved from Technology Corner

[time2roll]

CA Traveler wrote:
What’s your take on this data?

While shade cuts into production if series or parallel.... you will still have power and get a battery charged with a series set up.

Impossible to properly compare to parallel unless you switch to 24 volt battery.

[CA_Traveler]

Sal,

I believe you were looking at the red line which is voltage with the right hand side axis. The amps at noon are 15A.

But as I'm sure you know parallel would provide more amps for full panel shade. In the case above the 6 diodes in the 2 panels would add about a 3V drop.

The bypass diodes can provide more power for partial panel shading. For example the time frame from 8:18-10:18 would have no power for parallel panels since there is shade on all 3 panels. The voltage stepping from 12:58 to 1:38 also shows effect of the diodes with partial shading.

Granted this example is panel end shade and mostly causes all 6 bypass diodes to conduct until the shadows begin to exit the front panels at 12:58. Whereas side shade would allow 1/3 or 2/3 of the panel producing power for a longer period of time.

[Salvo]

Around noon output is about 9A. In full sun, one panel should output 35A/3 = 11.6A.

Would you be better off with parallel where one panel in sun and 2 in shade will give 11.6A?

BTW, there's two types of shade. Bypass diodes will conduct when shade is produced from a nearby object. Bypass diodes will not conduct when the shade is from an object that's further away. In this case the shade is not as hard and the shaded cells are producing a small amount of current.

[CA_Traveler]

True enough but this was done to gain a better understanding of bypass diode operations which has been posted about many times. So here is a real world field observation.

[EsoxLucius]

If my batteries were charged that day I wouldn't give it another thought.