Edge of Cloud Event

Roderick · March 2007

This was brought up as a tangent of another thread, but I thought the phenomenon should have a thread of its own.

The observation is that when the sun has been behind a cloud, and just peeks out, there is a spike in power output from the photovoltaic array. I've noticed this, but haven't studied it in detail.

I've heard (internet legends) that

1) This happens because the edge of the cloud focuses the sun.

2) This happens because the array has cooled down while the cloud passed over.

3) Something to do with a burst of new carriers (electrons), regardless of the temperature.

4) Because most of the sky is white clouds at the time, so besides the direct sun, light bouncing back from the other clouds hits the array.

5) It's just generally colder on a cloudy day.

Opinions, anyone? Sources to cite?

Does this effect last only a few seconds, or until the sun goes behind the next cloud? That piece of information would help sort between the theories above.

mike95490 · March 2007

Re: Edge of Cloud Event

> 2) This happens because the array has cooled down while the cloud passed over.
> 3) Something to do with a burst of new carriers (electrons), regardless of the temperature.

2 & 3 get my vote, I'd suspect 2 much more than 3. Mike

halfcrazy · March 2007

Re: Edge of Cloud Event

I believe i read somewhere that it was #1 the cloud edge actually magnifys the sun

crewzer · March 2007

Re: Edge of Cloud Event

It's basically (1) X (2). The focused sunlight increases current output, and the relatively cool module's Voc and Vmp are relatively high. The effect passes quickly as insolation retuns to "normal" and the PV cells heat back up in the direct sunlight.

HTH,
Jim / crewzer

System2 · March 2007

Re: Edge of Cloud Event

My experience is something else.
When you have a clear blue sky, the only source of light is directly from the sun. The rest of the sky is a relatively dark blue.
Now, put a bunch of bright clouds in the sky and space them out so the sun can shine full power between them. Now you have that same full brightness of the sun, BUT IN ADDITION, you now have the very bright edges of the clouds adding extra light from the sky.
Try looking directly at those bright edges and you'll find it hurts your eyes just like trying to look at the sun. Not as much, but it still hurts. Now hide the sun behind one of those clouds and your PV's still put out, using only the light from the edges of those bright clouds.
Sorry, but I thought this was common knowledge. I've seen it many times and the extra output from the PV's is rather dramatic and lasts as long as the sun shines directly on the PV's, along with the extra light from the bright clouds, both at the same time. Take away either the direct sunshine, OR the bright clouds and the PV output returns to "normal". It's a function of light intensity, very similar to full sun, ALONG WITH additional light reflected off snow covered ground, assuming the PV's are properly tilted for winter use in the higher latitudes.
Wayne

System2 · March 2007

Re: Edge of Cloud Event

Thanks everyone for your resoponses. The event is something I've seen all the time, just not thought about critically. So clouds aren't all bad to solar power producers, just needed in moderation.

I'd always thought the peaks I'd find on my MX60's logging function were a result of increasing loads when maximum power was going in at high noon, now I realize it's a combo of loads and edge of cloud events. 1750 watts has produced almost 2300watts once in a while, clear cold day with a few fluffy clouds...suprise surprise.

ralph

mike95490 · March 2007

Re: Edge of Cloud Event

As often as I've seen cloud shadows on the ground, either walking or driving, I've never noticed a bright "focused" spot at the edge. Is this an "every cloud" occurrence, or only a once in a great while event? I've seen my array output spike way up on partly cloudy days, but never see the corresponding bright spots at the shadow edges, so I'm assuming they are solely related to the panel cooling.
Are there any meteorological sites that discuss this phenomenon ? I've got an open mind, but I just have not see the bright spots, except in PV related postings.

cloud lensing: turns up astronomy sites
water vapor lensing : same results [Hubble Space Telescope Image of Gravitational Lensing. ... Water vapor absorbs infrared light.]
cloud edge effect: some suggestions that it only affects the IR spectrum
cloud focus: na-da

crewzer · March 2007

Re: Edge of Cloud Event

I think we're talking about variations of the same effect. Whether it's a "passing cloud" type or a "hole in the clouds" type (and I have to admit that I hadn't thought about that "alignment"), the water vapor in the edge of the cloud will focus (refract?) some sunlight into an intensity greater than "normal" insolation -- a natural magnifying glass, if you will. The result of this intense insolation will be greater current output from the PV module.

The effect is magnified if the module is relatively cool when struck by the focused sunlight because the Voc and Vmp will be relatively high until the cells heat up.

The effect is usually brief but quite real, and, as noted previously, it's a key reason what the NEC requires system surge capacity. For example, although my MX60 is rated for 60 A (continuous) in, all the NEC rules allow is for 48 A Isc in, which provides for a 25% surge capacity.

HTH,
Jim / crewzer

SolarJohn · April 2007

Re: Edge of Cloud Event

This is interesting. It explains the voltage spikes that my data logger sometimes records. Does anyone have any additional information on the subject?

John

crewzer · April 2007

Re: Edge of Cloud Event

John,

You may find the the results of this Google search string to be useful.

HTH,
Jim / crewzer

boB · April 2007

Re: Edge of Cloud Event

I understand this phenomena to be a "prismatic" effect. So, number 1 gets my vote. Edge of cloud
would be a magnifier effect. Temperature of a PV module can't change ~too~ fast because of its
size I wouldn't think, and I seem to remember that the current rise during the onset of edge of cloud
happens very quickly. Is that your observations too ?? Might also have to do with some filtering
effect of the EOC and applying a more responsive part of the spectrum to the PV ??

boB

crewzer · April 2007

Re: Edge of Cloud Event

boB!

There’s no doubt about #1, in my view.

Re the module temperature, I agree that PV module temperature change is fairly slow during the “onset” of EOC – passive radiation? However, it’s the re-emergence of the Sun from behind the passing cloud that causes the voltage spike when the sunlight light hits the “cool” module (#2). In this case, the module then tends to heat up fairly quickly (~80% of the insolation energy is converted to heat), which of course then causes the module's instantaneous Voc and Vmp values to drop. The behavior lasts for several seconds and is observable with a DVM or the meter on a PWM controller such as the Morningstar… I haven’t tried it with certain brand “O” controllers… :evil:

So, I continue to maintain that the spike in power output is a combination of (1) and (2), although the variations in voltage and current aren’t exactly synchronized. Another twist to this is that the battery voltage will rise briefly during the current surge – a variant of “surface charge”.

I like the chart on this page (scroll down a bit): http://www.daycreek.com/dc/HTML/journal061504.htm

Regards,
Jim / crewzer

4/20/07 Add-in Edit: In thinking about this issue during lunch today, it seems to me that the MX60 controller, operating in BULK mode, might be a good tool for observing details of an “edge of cloud” event.

Specifically, the front panel display presents both PV input voltage (measured) and PV input current (calculated, as I recall). If just the array current is affected by an EOC event, the PV current value should increase dramtically for a short time, and then return to a "normal" value. If, however, the PV voltage is also affected, the display PV voltage value might surge up and then down as well, depending on when the last MPPT mini-sweep was conducted (under cloud cover or under Sun?) and what MPPT PV voltage (PWM %) was set.

I think… :?

These changes should be independent of any changes in the battery voltage and/or charge current, as they’re on the “upstream” side of the MX60’s DC-DC buck converter. One thing to bear in mind is that these power spikes are brief in duration, so we're not talking about a lot of energy (power x time) here.

We’re supposed to have full Sun for the remainder of today and all weekend, so I don’t think I’ll have an opportunity to experiment… :-(

SolarJohn · April 2007

Re: Edge of Cloud Event

Thanks everyone for your response to this. When time permits, I'm going to perform a simple test. With my data logger set up to take samples every one second, I'm going to cover a solar panel and let it cool down. Then, in full sun with no clouds, I'll remove the cover. If I see an abnormal spike (before the panel warms up), then #2 will have been proven to be at least partially responsible for the effect. Has anyone else ever tried this, or any other tests?

John

crewzer · April 2007

Re: Edge of Cloud Event

John,

I think you'll find this to be an interesting experiment. It'll probably be helpful to keep the cover a couple of inches above the module to allow it to cool properly. I also suggest you consider conducting your experiment in two ways:

1) With the PVC module connected to a controller and battery, and note the controller's operational state (bulk, absorb, or float)
2) With the PV module disconnected from your system, and operating "open circuit".

Looking forward to your results!
Jim / crewzer

boB · April 2007

Re: Edge of Cloud Event

Great idea !

Also, notice the time and drop-off in output as the panel heats up more.
And how much of course the power increased when you pulled off the cover.
Is it as much as an edge of cloud event ??

Somehow I am guessing that most of the increase in output is because of the
focusing effect of the eoc and the raising of the Watts/mtr^2.

Make sure the cover is white (for reflection) I guess and if possible, raise it so it's not quite
touching the panel. This should let it cool a bit better while it's covering it up I think.
Maybe use a couple of saw horses ?? Are these panels on the ground or on a rack ?? It might
not be too easy to keep the cover off the panels.

You could also take a hose and spray water on the module and watch the power go up.
Be careful you don't crack it though. I haven't had any problem with it, but have heard that
some panels can have a problem with too cold of water. A light spray first should be better I
would think.

boB

fatheadl · April 2007

Re: Edge of Cloud Event

Hi all,

I have 2 KC130 in series, and on couldy days i have seen max power climb to as much as 315 watts. The mx60 display shows increase in panel current 10+ amps, do not remember a change in panel voltage.

fatheadl

mike95490 · April 2007

Re: Edge of Cloud Event

SolarJohn wrote:

Thanks everyone for your response to this. When time permits, I'm going to perform a simple test. With my data logger set up to take samples every one second, I'm going to cover a solar panel and let it cool down. Then, in full sun with no clouds, I'll remove the cover. If I see an abnormal spike (before the panel warms up), then #2 will have been proven to be at least partially responsible for the effect. Has anyone else ever tried this, or any other tests?

I think one of those $10 sheets of foil backed rigid insulating foam from a big box building supply store would work well as a cover.

niel · April 2007

Re: Edge of Cloud Event

quoted from boB:
"You could also take a hose and spray water on the module and watch the power go up.
Be careful you don't crack it though. I haven't had any problem with it, but have heard that
some panels can have a problem with too cold of water."

it seems to me that the pv that can't take cold hose water isn't going to survive long with mother nature. intense storms that brew up on a hot summers day put the cold water on the pvs along with hail at times after being superheated in the sun. hmmm, that could be sanyo's next excuse for their warranty denials. uh, the hail didn't break the glass, but the cold did it. warranty denied. if you know of pvs that don't survive cold hose water i'm asking you to reveal what pvs they are.

BB. · April 2007

Re: Edge of Cloud Event

My concern with cold water tended more towards the issue of a sudden vacuum drawing in water through the electrical box, connections, and glass/cell seals. I have double pane windows which I can draw in water if I drench them on a hot day.

-Bill

boB · April 2007

Re: Edge of Cloud Event

>>>hmmm, that could be sanyo's next excuse for their warranty denials.

Yeah, that's probably about right !

I'm talking about a very sudden large change in temperature, but I don't really know the reality of this since
whenever I've sprayed the hose on PV I didn't have a problem.

Delamination was the term I had heard used with this problem.. Evidently this was worse in older modules
and is worse in hot and humid environments.

Here's a quick description statement from one web site:

Module Delamination
A common failure mode in early generations of modules, module delamination is now less of a problem. It is usually caused by reductions in bond strength, either environmentally induced by moisture or photothermal aging and stress which is induced by differential thermal and humidity expansion.

And here is a decent article from Sandia and the FSEC that discusses delamination much better:

http://www.sandia.gov/pv/docs/PDF/prmking.pdf Starts on around page 6.

I have never seen a problem with using the hose but I originally heard about the problem from an experienced installer
in Hawaii.

boB

niel · April 2007

Re: Edge of Cloud Event

that's very interesting and that i actually may be on target about sanyo. :-o now it would be better i'd think to do this and see if the modules fail in their first year by doing this very thing unless they are pulling what sanyo is pulling.

fatheadl · April 2007

Re: Edge of Cloud Event

Hi all,

Some things i observed today at 12 o'clock EST. It was very cloudy (cloud cover 85% Accuweather says) and temp about 31C.

PV volts 33 watts up to 300
sweep (sweep interval 3 minutes)
PV volts 29 watts up to 290
sweep
PV volt 31 watts up to 280
sweep
PV volts 19 watts up to 260
sweep and autostart
PV volts 31 watts up to 290

Before I wrote this i checked log2 and highwatts was 310

fatheadl

crewzer · April 2007

Re: Edge of Cloud Event

PV volts 19 watts up to 260

fatheadl,

That was a substantial drop in voltage (~35%), but only ~ a 10% drop in power. Were you able to correlate atmospheric conditions to this event?

Or, conversely, what caused the relatively large voltage "spikes" (from 19 V up to 29 V - 33 V, or +52% to +74%) accompanied by smaller power gains (up from 260 W to 280 W - 300 W, or +8% to +15%)?

Thanks,
Jim / crewzer

fatheadl · April 2007

Re: Edge of Cloud Event

Hi Jim,

The panels are on a concrete roof 2 floors up so i could not get to the panels to check. I am assuming the panels got very hot.
From what i saw there was no change in voltage except after a sweep. The current was 10.xx amps The 19v did shock me.

fatheadl

BB. · April 2007

Re: Edge of Cloud Event

If I understand correctly... Many of the MPPT controllers for DC (as opposed to Grid Tie) sweep the current to find the optimal voltage for power generation every 5-15 minutes (don't remember which). So--once a controller like the MX 60 has found the optimal voltage for the next period, regardless of sun or temperature, the voltage will be kept relatively constant while the current is adjusted up or down to keep the voltage the same.

A Grid Tie Inverter (like the Xantrex GT series), per Solar Guppy one of the designers, constantly adjusts for maximum power (multiple times per second or even per cycle???).

-Bill

boB · April 2007

Re: Edge of Cloud Event

>>A Grid Tie Inverter (like the Xantrex GT series), per Solar Guppy one of the designers, constantly adjusts for maximum power (multiple times per second or even per cycle???).

Which means that it rarely spends a lot of time sitting at the MP point. It's always looking around. This is one big tradeoff
of course. It probably doesn't go very far off the MP Point though. Hopefully. BTW, Henry Cutler has a patent on his
and Xantrex' sweep technique.

boB

crewzer · April 2007

Re: Edge of Cloud Event

So--once a controller like the MX 60 has found the optimal voltage for the next period, regardless of sun or temperature, the voltage will be kept relatively constant while the current is adjusted up or down to keep the voltage the same.

boB,

Is that correct? I kinda thought the MX swept the array voltage looking for the optimal PWM duty cycle % that resulted in the highest charging current. If so, a change in array cell temperature after a "new" MPP is found could (would?) affect the array voltage. I know that I’ve seen my MX’ PV voltage change between sweeps…

I found the Cutler/Xantrex patent to be veddy interesting…

Regards,
Jim / crewzer

boB · April 2007

Re: Edge of Cloud Event

>>Is that correct? I kinda thought the MX swept the array voltage looking for the optimal PWM duty cycle % that resulted in the highest charging current.

Yep. Unless you "KNOW" the MPP Voltage, ya gotta find it somehow. And that "usually" means moving it around and looking.

Another way to sortof know what it is without sweeping the array or doing "something" is to sweep a reference
cell that isn't part of the main array so you don't waste any energy. Of course doing that isn't really finding
the max power point of the array that you're sucking off of.

Kind of a CooL idea though.

boB

crewzer · April 2007

Re: Edge of Cloud Event

Very interesting! OK; thx.

Regards,
Jim / crewzer

Edge of Cloud Event

Comments

Categories