Large drop in solar power in one year

emerika · June 2011

Last year I purchased 12 of the 210 watt Kyocera solar panels (KD210GX-LPU). When I first installed them (four groups of three in serial) I saw 88 volts and I routinely was able to produce 2600 watts (as high as 2800) of power. This summer our power has gone down a lot. Today there are no clouds. I am producing a maximum of 1700 watts and I see appox 69 volts from each group of three panels.

Things I have done to address this:

1) I have washed the panels with a sponge and water
2) readjusted the angle to the sun
3) Tightened all wiring
4) Checked that nothing is casting shadows on the panels

Things I have checked:

1) measured volts at each of the four groups (at the DC breakers) 69. volts (some what less than the 78 I expect; however, each is within a volt or two of the others.
2) measured the amps at each group: again, all groups are working about the same.

I did the measurements disconnected from the rest of the system.

Any thoughts? Is this just a normal drop-off after a year of use? 1700 from an expected 2600 is an awful lot.

You can look at the installation here: http://www.kbtechworks.com/kbranch/blog/?tag=solar

brad

BB. · June 2011

Re: Large drop in solar power in one year

If you do not have one, I would suggest getting a DC current clamp meter (like this $60 entry level version). This is the easiest and safest way to check DC current flow in a solar PV system (panels, battery bank cabling, etc.).

Check each parallel string of solar panels--Ideally, you want them to match current flow to within 10% of each other... If you have several producing a lot of current and one producing less than 1/2--Then you may have bad electrical connections or a bad panel.

I am assuming that these panels are operating on a XC 60 Amp MPPT charge controller?

Next, check the solar panels for brown/black spots (electrical heating). It may not be obvious--and on my panels, I have trees leaves that sometimes make a "tea" in the lower edges of the panels--a dark stain on top of the glass that looks very much like a "charred" panel. Washing the low spots with soap and water should remove any staining and let you see the panels better.

On my array (different brand), I had them go from nearly 100% of the power that I expected to 50% of what I expected (Grid Tied System) in a matter of a week or so.

And when I looked at my known bad panels on the ground (installer would test each panel with a DMM to measure Voc and Isc)--I at first though they looked fine.

However, my installer showed me the problem they saw with that vendor's panels. On the glass side of the sealed junction box, there was a light coffee/rust colored stain above the box. And on some of the panels, I could see lamination (streaks of bubbles) around the outline of the box--Probably signs of overheated junction box.

With my vendor, if >20% of the panels were "bad"--they would replace the whole array (all my panels were replaced). Kyocera has a good reputation on warranty service from about 10 years ago when they had some failed panels.

Normally, solar panels are probably the most reliable portion of the system. So, I would suspect failing wiring connections (loose, corrosion, bad crimps, etc.) first. Then I would suspect the MPPT charge controller, and lastly the solar panels themselves.

Obviously, the state of charge of your battery bank matters too, as does battery wiring (again, checking current flow among parallel battery banks with the DC current clamp meter). If the batteries are >80-90% charged, they may not take full power. If there is a bad connection in a parallel bank--then that section of battery bank is not charging/discharging. Using the DC current Clamp Meter to measure current flow in parallel paths (under heavy charging or heavy load) will help you verify that all is OK (or not) with battery bank wiring/cells.

However--from your description--I would look at the solar panels very closely.

Use a DC current clamp meter to check each string for current flow.
Use a DMM to measure the Voc of each panel (open circuit voltage)
Use a DMM set to 10 amp scale and measure Isc in full sun (if your current is close to Imp/Isc--all is good. If 1/2 or less, probably a bad panel.

Before you disconnect panel wiring, make sure the DC disconnect is open (or do work in "dark". The MC4 connectors do not like being disconnected under load.

Anyway--that is how I would start.

Good Luck!
-Bill

Vic · June 2011

Re: Large drop in solar power in one year

BB Bill, as usual, has some great tips.

AND, I am not certain that there is a problem with your system, altho, you did note a difference in one year, perhaps 12 months, or perhaps 2010 into 2011.

Think that you are in the desert SW, where it just might be HOT. In comparing panel output between cold outside temps and HOT, could explain the difference. The Vmp of PVs are reduced with rising temps, as is the current available.

I know little of your system, other than prob off grid and 12 ea Kyo 210s, but the STC for your array would seem to be about 2520 Watts. It is always good to gently clean the PVs, check connections, as you have done, and BB Bill has some great tips. The clamp-on DC ammeter is also a great tool.

As always, whatta I know ?! Vic

Derik · June 2011

Re: Large drop in solar power in one year

I noticed my power output from the panels was low when I checked my log on the charge controller. I observed it for a weekend and what I noticed was that my system was taking in sun so early and reaching float so early in the day that the panel output was lower than normal. On days where I really used the power or where it was cloudy and the bulk setting was reached later in the day then the panel output would be back up again.

So in other words my panels were in bulk by 10:00 a.m. prior to the sun being high enough for the panels to make maximum power. By noon the voltage was maintained in bulk with out needing too many watts from the panels and by 1:00 I was in float.

Hope that makes sense.

Cariboocoot · June 2011

Re: Large drop in solar power in one year

You can't judge off-grid harvest this way. Unlike grid-tie where you see everything the panels put out, off-grid is dependent on how much the batteries need. If they haven't been drawn down the same as "last time" they won't need the same amount of charge. Once charged, at whatever rate or time, the panel production falls off because there's no place for it to go.

Is the system still providing the power you need every day? If so, don't worry about it.

Dave Angelini · June 2011

Re: Large drop in solar power in one year

Derick,

Obtain a small 1500 watt heater and load test your system. The xwcc logs will only be the peak for that day and conditions are not repeatable. It is helpful to watch the logs but a load test is much more accurate. Set the heater by your SCP and just observe the system at different times. It only needs to be on for a minute. Make sure the system can maintain the power as loose wiring often is revealed by this test.

Beautiful place! You guys are fortuanate! Take care!

emerika · June 2011

Re: Large drop in solar power in one year

I should have been clear that with my batteries only 65% charged I was only seeing 1800ish watts.

I am familiar with the power from the solar panels dropping once the charge goes into float. This is not the case. To be sure I turned on a 1500 heater. No change in PV watts.

I ordered a a DC current clamp meter and continue the search.

Thanks for the helpful comments.

techntrek · June 2011

Re: Large drop in solar power in one year

emerika - looking at your revised battery wiring diagram at the very bottom of your page, you need to revise your wiring again. Bottom line, the length of wire used to connect each string (+ and - added together) needs to be equal for all strings. Take a look at this page for info:

http://www.smartgauge.co.uk/batt_con.html

Here is a link to the diagram on your site, I didn't embed it to cut down traffic to your server : http://www.kbtechworks.com/kbranch/blog/wp-content/uploads/2009/12/batteriesfinal-300x231.png

Edit, now I'm not sure, I started counting the "longs" and "shorts" used on the the 3 paths available for + and the 3 available for each - , and maybe overall they are equal length. I started getting a headache trying to keep the total counts straight and gave up... gotta get some real work done at work! Maybe someone else has the patience...

Cariboocoot · June 2011

Re: Large drop in solar power in one year

Egad. Six parallel banks on a 48 Volt system. That's asking for trouble.
When you need that much capacity it's time to think about 2 Volt cells or forklift batteries.

BB. · June 2011

Re: Large drop in solar power in one year

Your wiring per techntrek's link--Not ideal. You can use the DC Clamp meter to check the major links to ensure each battery string is sharing current (but with parallel paths--it is not easy to say what the "proper current" each path should have).

Also--check the individual battery voltages and make sure that they "match" (any very low or very high battery voltage can indicate problems). Also, check the "48 volt" battery bus terminals too (they should all be virtually the same voltage).

However, I would be surprised if your battery system is responsible for any "loss of solar power" you are seeing.

-Bill

techntrek · June 2011

Re: Large drop in solar power in one year

Making a new post here to point out I edited my post above - now I'm not sure that there is an imbalance but taking measurements as Bill suggested is the only way to be sure. There are so many connections and long/short wires that estimating it on paper doesn't mean much. As I say above, its possible they really are equal, but I got a headache trying to count everything up!

I agree that this doesn't have anything to do with your original question about the PV output, but still something you need to look at. I think you can simplify your battery wiring a lot - which will reduce the number of failure points.

emerika · June 2011

Re: Large drop in solar power in one year

techntrek wrote: »

Making a new post here to point out I edited my post above - now I'm not sure that there is an imbalance but taking measurements as Bill suggested is the only way to be sure. There are so many connections and long/short wires that estimating it on paper doesn't mean much. As I say above, its possible they really are equal, but I got a headache trying to count everything up!

I agree that this doesn't have anything to do with your original question about the PV output, but still something you need to look at. I think you can simplify your battery wiring a lot - which will reduce the number of failure points.

Thanks a ton for looking at the battery wiring. I was lead to believe that though complicated, it would be better to have redundant wiring. It sounds like it's not the case.

I will look at the information you provided and rethink the battery wiring and look to a cleaner solution.My wife is out of town next week so I can make changes then.

Back to solar panels.

Dave Angelini · June 2011

Re: Large drop in solar power in one year

All of your solar breakers/fuses/connections are clean and torqued right?

emerika · June 2011

Re: Large drop in solar power in one year

Maybe Cariboocoot is right and you just can't measure an off-grid system. I don't know.

I've cleaned the panels (again) with a soft sponge and water, I fine tuned the angle to the sun. It's high noon and I am only seeing 2000 watts. A small improvement.

Still, it looks like the batteries will be fully charged in the next hour. By 1:00 PM

It is hot here. I know this affects the charging of the batteries. Perhaps the SCC is not using all the power that is available. ??? Still I just ran a 1500 watt heater and the power from the solar panels didn't go up. All that happened is that less power went into the batteries.

I'm still waiting for the DC current clamp meter. I think that when it comes I will disconnect the solar array at night and wait to turn it on until there is good sun. This way the batteries are in strong need of a charge, the panels will be in position to deliver a lot of power and the SCC should want all the power it can get. Then I can use the meter and see what is what.

The fact that disconnecting any of the four blocks of solar panels always results in 1/4 power loss to the SCC suggests the SCC is using all the power available.

Batteries - I see that my final diagram was not what I actually did. Here's a corrected diagram which hides the 12 volt batteries and only shows the 48 volt batteries.

http://www.emerika.com/batteries_actual.png

Looking at: http://www.smartgauge.co.uk/batt_con.html, I could do easily do method 2. Method 3: I'm not doing because I'd have to buy new cables. As for method 4, if I could find a diagram for 6 batteries I would do method 4. (anyone)

I am curious though why the redundancy I've added doesn't just even everything out. Of course, since it's not obvious to me why method 1 is worse than method 2, I'm not the person to ask.

The system's been working for over a year now. I still have much to learn. Thanks to everyone taking the time to lend a hand here.

Cariboocoot · June 2011

Re: Large drop in solar power in one year

You could always disconnect all the panels and check each one's Isc about midday and see if any are out of spec. It's tedious, dangerous, hot, boring, annoying ..... But it will tell you if the panels are okay or if any of them have gone bad.

If some of that massive battery bank has become sulphated or disconnected (even partially) from the system the panels will not have as much charging to do. Long shot, but it's there.

mike95490 · June 2011

Re: Large drop in solar power in one year

Another ez trial, disconnect PV at night, and wait till solar noon to reconnect, and measure quickly.

Vic · June 2011

Re: Large drop in solar power in one year

In addition to all of the above,

There are factors that govern the irradiance, such as water vapor, smoke, dust and other particles in the air. These are variables, as are all of the weather and temperature factors. These make it more difficult to judge one day/week .. year to another, and are harder to see than clouds, or the thermometer

Seems to me, that in the desert SW that, 2000 watts on a hot summer midday is just fine from a 2500-ish watt array. Perhaps my standards are just too low. My panels are always at the winter tilt angle, and am happy to get 1900 watts midday from 3.15 Kw ... may need to change the panel angle, it is HOT here, too.

Have Fun, Vic

solar_dave · June 2011

Re: Large drop in solar power in one year

Vic wrote: »

BB Bill, as usual, has some great tips.

AND, I am not certain that there is a problem with your system, altho, you did note a difference in one year, perhaps 12 months, or perhaps 2010 into 2011.

Think that you are in the desert SW, where it just might be HOT. In comparing panel output between cold outside temps and HOT, could explain the difference. The Vmp of PVs are reduced with rising temps, as is the current available.

I know little of your system, other than prob off grid and 12 ea Kyo 210s, but the STC for your array would seem to be about 2520 Watts. It is always good to gently clean the PVs, check connections, as you have done, and BB Bill has some great tips. The clamp-on DC ammeter is also a great tool.

As always, whatta I know ?! Vic

Yeah my output hits max in late April/ early May even though the days are much longer and the angles are pretty good, the TEMPS are Way HIGH. Temps today are 110+F and with an infrared gun the backs of my panels are 145+F this afternoon. Production is off about 15% from my spring noon time peak output.

Dave Angelini · June 2011

Re: Large drop in solar power in one year

On your input menu the "auto tracking should be enabled". You can see reference voltage change with temperature.

emerika · June 2011

Re: Large drop in solar power in one year

I found this on another forum.

For every degree C above 25° C you will lose approximately 0.5% production.

It would not surprise me to find the temperature at the panels in the 130 degree range if not hotter. Which reduces the maximum output of a panel to something 178 watts. This explains enough of the power loss that I will attribute the rest to noise.

A somewhat interesting phenomenon for the south west. The days are so hot in the summer, but the loss in panel efficiency is offset because the days are longer. In the winter when the days are short, the panels are more efficient.

ggunn · June 2011

Re: Large drop in solar power in one year

emerika wrote: »

I found this on another forum.

For every degree C above 25° C you will lose approximately 0.5% production.

Usually it's closer to 0.35%, but there is some variation. The data sheet for your module will usually show a temperature coefficient for voltage expressed as either a percentage or millivolt drop per degree C. There is also a current gain for elevated temperature, but it is much smaller than the voltage drop, so to a first approximation, the power tracks with the voltage.

BB. · June 2011

Re: Large drop in solar power in one year

I think you still may have real problems. That is a significant drop... Especially the voltage drop.

Check for poor current flow/overheated connections.

If you do have a real voltage drop, there is a good change that the problem will get worse. So keep an eye on the system.

It is a close call at this time.

-Bill

emerika · June 2011

Re: Large drop in solar power in one year

BB. wrote: »

I think you still may have real problems. That is a significant drop... Especially the voltage drop.

Check for poor current flow/overheated connections.

If you do have a real voltage drop, there is a good change that the problem will get worse. So keep an eye on the system.

It is a close call at this time.

-Bill

Thanks - I will keep looking. I received a clamp meter too late this afternoon to do any examination, but I will tomorrow.

Tonight I'll read up on using the clamp meter. The one I got has volts for AC and DC, but it only shows A as AC. I fear I got the wrong thing. I'll start reading first.

I did however, check the volts for each of the batteries: both the 12 volts ones and the 48 volt ones I made up. In every case all batteries had the same volts: 12.6 and 50.1 respectively.

BB. · June 2011

Re: Large drop in solar power in one year

Yikes! Sounds like the wrong meter. There should be both a DC and an AC Amps scale/setting.

12.6 volts / 50.1 sounds like the batteries are not being charged--and may either have a little load on them, or not not 100% charged (also depends on battery temperature). But they are not getting damaged at the moment.

-Bill

Dave Angelini · June 2011

Re: Large drop in solar power in one year

Too bad about the meter. Along that line, I never read that you have in fact torqued all the connections, breakers, and inspected any butt splices. Most of the time it is the connections and not the panels.

Not a bad time though to get panel serial numbers and contact the Rep you bought them from. Might be a recall. That is too much power to loose that fast, no matter what temperature!

emerika · June 2011

Re: Large drop in solar power in one year

Mysteriouser and mysteriouser- Maybe it is the heat. Today has been cool: 99 degrees. Also a little cloudy. During testing with a 2000 watt load I suddenly noticed I was seeing a little better than 2300 watts. Frankly it's pretty much been 110 here for awhile.

1900 watts = 73 percent of maximum
2300 watts = 88 percent of maximum (I'm oddly comfortable with this number)

Backing up though, I got the correct current meter and as I suspected each of the four groups of panels shows the exact same numbers. There's no reason to think any one panel is bad. Either they all are good or all bad.

I'm going to look through the log files and see if I can show a correlation between maximum output from the panels and temperature. It's a bit tenuous, because the maximum may not have been reached on a particular day for more then one reason. Still, it's likely a relation can be seen.

What does anyone think about water cooling the panels to see if cooling them down temporarily show an immediate increase in power?

I emailed both Kyocera and Xantrex asking for a little technical assistance and have heard nothing but silence.

BB. · June 2011

Re: Large drop in solar power in one year

Those are reasonable numbers and I would not worry. It is possible that the other stuff you did (checking connections) may have helped too.

I did have distributed failures in my solar array--so measuring the two different strings really did not show much.

On my GT array (always 100% of available sun--no batteries to charge)--I never saw a marked difference between hot and water cooled panels (probably less than 10% change in output levels--my guess).

Personally, I do not like th thermal shock stuff (cold water on hot panels, etc.)--I think it is hard on the devices and can affect long term life (differential compression and expansion between cover glass, cells, and backing).

However--One of the guys here with a whole bunch of solar equipment experience (Solar Guppy) has never seen de-lamination/other problems from cold water on hot panels--So it does not appear to be a big issue either way.

-Bill

Derik · July 2011

Re: Large drop in solar power in one year

I think your fine, I ran a small ac unit yesterday to see if I could get my 9-215 watt panels up. It was about 100 degrees outside, panel temp on the top side was 124 degrees and all I could get was 1570 watts or so of power.

I don't remember the summer production being that low last year but then again, it simply never got this hot last year.

I heard the prediction for the SW is a short hot summer.

emerika · July 2011

Re: Large drop in solar power in one year

Derik wrote: »

I think your fine, I ran a small ac unit yesterday to see if I could get my 9-215 watt panels up. It was about 100 degrees outside, panel temp on the top side was 124 degrees and all I could get was 1570 watts or so of power.

I don't remember the summer production being that low last year but then again, it simply never got this hot last year.

I heard the prediction for the SW is a short hot summer.

Thank for chiming in. That sounds a lot like me alright. We've had a few cloudy days here which is providing interesting data. When it clouds up for awhile and then gets sunny, I see pretty normal number for awhile. I guess the panels cool off (somewhat) being out of the sun for awhile. I saw 2500 watts for awhile the day before yesterday. Today the max was 1900.

It's nice to know... or at least have an idea what's going on.

emerika · August 2011

Re: Large drop in solar power in one year

Just a follow up to my own post. Things have cooled down a bit and I am again routinely seeing 2600 watts (or better) of power being generated. It was the heat - end of story.

Large drop in solar power in one year

Comments

Categories