Solar, And Temperature

These cool, partly cloudy spring days provided me an opportunity to see the temperature dependence of solar panel energy generation right up close and personally.

In one particular instance, it had been cloudy for at least an hour.  There was a gentle breeze blowing and the temperature was in the high 40s.  Because of the breeze and the shade, my solar panels were cool.

Then the clouds parted and the sun came out, pretty much all at once.

551.6 watts

I know it's a little hard to read, but the blue system monitor at the lower right is showing 551.6 watts.

A few minutes later, the solar panels were still in full sun - that is, no change in energy falling on the panels.  Yet, here is the output:

Later, 532.4 watts

Yup, output has dropped to 532.4 watts as the panels heated up...  that's a drop of 3%.  That is a pretty significant drop.  You folks in the tropics should be able to get a pretty substantial increase in solar output if you can figure out some way too cool your panels...

(FWIW, the biggest reading I saw came later in the day when the sun was higher, and again just after a cloud has passed:  571 watts.  I love seeing that current meter needle buried!)

Solar, And Temperature

Another Ten-year Project

Well, almost ten.

Way back in 2010, I bought a pile of 6" square solar cells, with the intention of making a bunch of solar panels.  I had no idea that it would take me this long to finish this effort.  This was a low-priority project, dependent on acquisition of used shower doors for free.  Thanks to craigslist, I was able to find the necessary shower doors, but this took way longer than I anticipated.

Nevertheless, finished it is.  By the time I had built 5 panels, the remaining cells I had were not enough to build that sixth panel, due to breakage in transit, breakage due to my clumsiness (these things are *fragile* - in comparison, a potato chip is way robust), and missing collector busses.  The seller had included extra cells to cover breakage in transit, but sadly none to cover for my clumsiness.  I might be able to resurrect those cells with missing collector busses, but I don't know if that would give me the required 40 cells.  So for now, I am finished.

For each panel, no-load output voltage (at zero current) is about 22V, and short circuit current approaches 8 amps, as advertised.  But as I have mentioned earlier, you don't get to have short circuit current and no-load circuit voltage at the same time.

Something else I have learned is that the output of solar cells is temperature dependent.  The cells produce significantly more power when they are cool than when they are hot.  And that dark color means that they will be hot in the sun.  The highest output I have seen from my array is on a day that is mostly cloudy (panels are shaded), when the sun breaks thru (panels are in direct sun, but still cool).

So, a realistic assessment of the power output from these panels is about 100 watts each, at our lattitude.  This means that I have the capability to produce a little more than 500 watts (including the three little panels that came with Eolian when we got her).  That power is directed into a grid-tie inverter...  there are no batteries in our system.  In essence I am using the power grid as my battery bank.  The inverter turns the output of the panels into 110V, 60 Hz and pushes it back into the line, synced with the line power (if the line power disappears, say during a storm, the inverter automatically shuts off to prevent back-feeding the line).  If my home is drawing more than 500 watts, part of that draw is supplied by the solar panels.  On the other hand, if the house is drawing less than 500 watts, then the solar panel system runs my electric meter backwards.  (With a maximum output of 500 watts, I don't think I will ever need to worry about what happens if I end a month with a negative meter reading.)

Sadly, the day that Jane and I got the last panel up onto the roof of my shop was the first day of the fall rainy season.  I have yet to see what the finished system can deliver in sunshine.  Nevertheless, it feels good to tie the ribbons on another long term project.

(The remaining posts on this project can be found here.)

Another Ten-year Project

Big Bertha was too big

Adding Big Bertha to her three smaller sisters was apparently too much for the grid-tie inverter... See that burned spot on the circuit board?  That marks the death of the inverter.

So now, with the summer solstice approaching and the best solar insolation of the year, I have no way to put that energy to work.   And I am here because I believed that an inverter advertised as having a 250 watt capacity would be able to handle 175-200 watts. 

Silly me.

Big Bertha was too big

Big Bertha #1 Joins Her Little Sisters

There she is, up there on the roof of my shop.  I had to mount her far enough from the edge of the roof to allow a foot path to get at the mounts (unlike her little sisters, whose mounts are only on the tops and bottoms).  In fact, you might ask why the panels are tilted slightly but at the same slant as the roof...  It is purely a concession to the high winds we get here - when they are strongest they come from the left in the picture.  By keeping the panels parallel to the roof and tilted slightly into the wind I prevent the wind from getting under them and lifting them - I hope.  The little ones have survived a winter, so that is some validation anyway.  I'd get more power from the panels if I tilted them more vertical...  until the first big windstorm.

And the crew is really producing power now!

But I fell into a trap when I last reported on the project.  I simply multiplied the short circuit current by the open circuit voltage to get a power output.  But that's wrong.  You'd never get the open circuit voltage when the output is shorted, and the current is zero when measuring the open circuit voltage.  The reality is that both the current and voltage will be lower at maximum power output.

In fact, any solar panel power collector should have MPPT circuitry (my grid-tie inverter does).  MPPT stands for Maximum Power Point Tracking - it means that the collector continuously varies the current being drawn while watching the voltage and looking for that magical balance point where the maximum power is being produced.  And it is not a fixed point, which is why you need a controller with active circuitry.  It changes with the amount of sunlight falling on the panel, with temperature, and whether any of the cells are partially or completely shaded at the moment.

So, as things stand right now, Big Bertha is producing about 110 watts in bright sunlight, and her sisters produce about 75 watts.  I checked our electric meter when everything in the house and shop was quiescent.  Indeed, it was turning backward, ever so slowly.  I can't wait to get more panels made!  I should have enough cells for another 4 or 5 Big Berthas, depending on how many I break in the fabrication process.

Big Bertha #1 Joins Her Little Sisters

First light

That's what they call it when a new telescope first sees the sky...  I think it applies just as well to a solar panel.  This is the one you saw me building earlier. 

In the early morning light (impatient, yes I know), it was producing 5.5 amps at 24 volts.  That's 132 watts, actual.  Not calculated, not predicted - actual.

And there was a big fat spark when I connected up the ammeter - do you have any idea how satisfying that is?

First light

Solar panels, on the cheap

Some time back, I wrote about how the solar energy market was in a state of flux. And I may have mentioned that I bought a kilowatt's worth of 6"x6" solar cells...  for $350.  That price is still pretty much the same.  I am embarrassed to admit that that post was made more than three years ago.

Eight cells to go...

Well fast-forward those three years to the present day. Being retired has given me the time to get some of the "must do" projects off the list, leaving time for the "fun to do" items. And the solar panels are one of these.

Since I am doing this on the cheap, the first problem was finding something to serve as a substrate.  Tempered glass is expensive, and aluminum extrusions are not cheap either.  But guess what?  Shower doors, made from tempered glass and already rimmed with aluminum  extrusions are free, or nearly so.  And a regular shower door is just the right size to hold 40 of the 6"x6" cells in a 4 x 10 matrix, creating a panel that will produce a nominal 8 amps at 22V...  175 watts!  Not bad, for a $60 investment (so far).

And I have enough cells to make 6 of these panels.  Tho they will be too big and heavy to be suitable for a boat, they will work nicely on the roof of my shop where they will feed grid-tie inverters, allowing me to sell the power they produce back to the grid.

Solar panels, on the cheap

The gift of power

Power to the people!

No, wait.

Free power!  Yeah, that's the ticket.

Clean, free power from the sun!

Well, no, not quite.

First: clean.  Manufacturing silicon from silicon dioxide (quartz) consumes about 12 megawatt hours per ton of raw silicon produced.  Yeah, that's *mega*watt hours.  Then, the raw silicon must be refined to get rid of impurities - an exotic chemical process followed in the end with an energy-intensive electrolytic process for recovering the metallic silicon.  In all, about 200 megawatt hours is consumed in the production of a ton of silicon solar cells.  That's a lot of power - power which must be generated somewhere.

Next: free.  In the recent past, solar panels have sold for about $10/watt - that is, a panel with a rated capacity of 100 watts (in the tropics, at noon) would cost about $1000.  That's hardly free power.  If you work out the numbers, at our local power cost of $0.08/kwh, that 100 watt panel would be paid off in 85 years.

But something has been happening in the solar cell marketplace recently.  Solar panels can now be found for under $2/watt (17 year payback), and the bare cells are available for amazingly low prices.  And so just before Christmas, I bought a kilowatt's worth of solar cells and paid an amazing $350 for them - that's $0.35/watt - a payback period of only 3 years.  Those are numbers I can live with!

An early Christmas present!

The actual purchase was 250 cells, each of which is 6" x 6", and will deliver 8 amps at 0.5 - 0.6 volts.  They arrived in 4 boxes and were well protected with packing materials.

Here are a couple of the cells, showing front and back.  They are incredibly thin and fragile.  That ton of silicon mentioned above made a *lot* of these cells.

Now to actually construct the panels in a way that will last, the challenge will be to keep the cost from climbing back to that $2/watt for which commercial panels can be had.

Stay tuned.

The gift of power