two axis solar tracking system

[Joseph]

I had read somewhere that one axis tracking makes more sense because of an only small advantage for a more intricate system that can adjust up/down position as well as left/right. If i were doing one for myself, I would try just left/right.

[philba]

I'm curious as the often quoted phenomenon of "the brightest spot in the sky may not be where the sun is".
When would this occur, and how often? Anybody have tracking data on sun position vs. max power position?
Economics of a tracker is the concern here. How often is it justifiable to track max power when a simpler system that tracks just sun position will perform 99% of the time?

reflections from glass can do this. Imagine 2 reflections falling on the wrong LDR.

[Bigglez]

reflections from glass can do this. Imagine 2 reflections falling on the wrong LDR.

Good point. Wouldn't this be less likely if the
sensors had limited viewing angle (collimators
for example)?

Also, if the sensor found a secondary target
(window reflection) it would be tracked until that
reflection is removed by the relative movement of
the sun.

What does the tracker do if it looses the target?
How is 'night time' different from target loss?

Perhaps the tracker has a limited view of the
sky, so that it predicts the approximate location
of the sun, and uses tracking to counter seasonal
and time of day variance only?

Now we're back to predictive tracking - location
of the sun regardless of cloud cover, based only
on the viewing point on the earth, and the time
of day.

[philba]

Now we're back to predictive tracking - location
of the sun regardless of cloud cover, based only
on the viewing point on the earth, and the time
of day.

Yeah, I think this is the better way to go. Super simple field install and align. Especially if it includes a GPS.

[jwax]

Hey! Agreed! I see too many problems associated with a LDR max Power tracking system than is justified.
Point at where the sun is supposed to be at that time/day ("predictive tracking"), and have no worries.

[MrAl]

Hi MrAl. My question is how often is that elaborate seek-the-max system going to be necessary?
Traditional solar modules today have a fairly wide acceptance angle. They won't provide any significant difference in output if the max power spot is +/- 10º either way.
What's the gain of complex pinpoint accuracy, and in the case of cloud bounce, only occasionally occurring as well?

Hi,

Yes that +/- 10 degrees should be right because cos(10)=0.985
very nearly, and of course cos(-10)=0.985 also, so the conversion
is close to 100 percent even at angle of + and - ten degrees.

What is the gain of complex pinpoint accuracy? Well, it really depends
on the cloud cover. You'd either have to find some studies on this
(there must be some out there in this day and age) or else you would
have to do some experiments on your own. You could build a highly
accurate detection system with several small solar cells and have
a uC collect the data, then compare it to the actual array used for the
power, just for one example, over the course of a year or so.
Would be interesting.

[jwax]

Tell me if I'm belaboring this point, and I'll be quiet.
What circumstance in the daytime sky would cause a solar module to output more power by pointing anywhere but directly at the sun?
Do commercial trackers point at max power in the sky, or do they track the predicted sun position?
I hope to learn something here, and maybe save money by going the "predictive" route in an installation.

[MrAl]

Tell me if I'm belaboring this point, and I'll be quiet.
What circumstance in the daytime sky would cause a solar module to output more power by pointing anywhere but directly at the sun?
Do commercial trackers point at max power in the sky, or do they track the predicted sun position?
I hope to learn something here, and maybe save money by going the "predictive" route in an installation.

Hi again,


I have worked with very large solar arrays in the past (rated in kilowatts)
but most of the experimention i have done was more or less after the
fact, so i wasnt directly involved with noting these kinds of things first
hand. I had to do other things, like determine how to best tie the
system into the line, phase locking, even redesigning the whole dc to
ac control system, but unfortunately the most i have done with the
matters we are talking about here is how to best max power track,
which means controlling the system electronically (not mechanically)
so that the array power draw would keep it at the maximum possible
power output.

This means i dont have any first hand data on what you are asking,
but from what i have read about this in various places there are
times when the cloud gets directly in line with the sun (like a big umbrella)
and so that spot is darker than the surrounding sky. In energy
terms, the watts per square meter with a plane perpendicular to the
sun/local earth location line would be less than another choice of
angle, which would point at an angle other than directly at the sun.
Imagine having a very small solar panel in bright sun at 12 noon, pointed
directly at the sun, then placing your hand maybe two feet above the array
and so blocking the direct rays of light. The other parts of the sky could
look brighter, and perhaps provide more watts per square meter than
when pointed at the hand.
What i can not say is just how much difference this can possibly make,
again because i was never directly involved in that phase of the design
of these types of systems. All i can say is that i have read that it is
better to do this, but that i dont have any data at all to back that up
at this point.

This might not be too hard to prove or disprove however. Perhaps
manually pointing the array at the sun when there will be some cloud
cover that day, then waiting for that cloud cover, measure panel
output, tilt the array to point to anther location off axis, then measure
panel output again and compare readings. It might take a few
tries to get any useful results but that basic idea should show you what is
what.
Perhaps even synchronizing a video camera with a solar panel output
and manual control over the panel tilt angles. In place of vid camera
could be several small panels with outputs monitored and logged.
As the cloud cover approaches it would be interesting to look at the
way the data changes and after it's over compare all the data and
look for the highest possible energy output and note which position
brought that about. The small solar cells are not that expensive,
so perhaps a small basketball size (semi sphere) array of cells
where the cells are mounted on the semi sphere. This would provide
a multi angle measurement system without having to actually move
anything mechanically. After a few cloud covered days the results
should provide the data needed.

[Pius]

Dear Solar Pulse

I too have been interested in having a tracking solar collector. I built a prototype that would track light but it just sits on my desk. As such I can give you no technical advice.

However over the years I have collected a few articles that detail building such devices. Let me know if you are interested and I will gladly send you this information.

Take care good luck,

Pius

[SolarPulse]

Dear Solar Pulse

However over the years I have collected a few articles that detail building such devices. Let me know if you are interested and I will gladly send you this information.

Pius

yep that would be of great help to me. ud be able to directly email me them from here.

well guys, thanx again for all the posts. if the solar panel tilts +/-10 degrees the voltage drop is about couple of hundred milivolts, i checkd it with a protractor attached to a simple system i used. either way as you've already discussed 98% accuracy is good enough.

well commercial trackers use both the chronological(suns supposed to be position) method and sensor based method(highest intensity in the sky).

even i think, calculated tracking specially if a gps available would be the easiest and best way. but i just wanted to try use a sensor, something simple, to track. after reading the posts i want to try using an array of CDS cells fixed inside small black tubes(to give it the collimator effect) fixed at an angle to each other. so the uC will try to maintain the highest intensity at the middle cell. to avoid hunting, im thinking of using a small pin like thing in the middle which would cast a shadow. what do u guys think about this? not sure if its gonna work until i actually try it.

[MrAl]

Hi again,

What i think is this...

As i was saying in earlier posts, solar sensors and panels respond to light not only as the intensity itself but also as the angle of incidence.
What this means is that the output of a solar cell will be dependent on the angle it is placed relative to a point light source. The law
is simply cos(angle), where 'angle' is the angle measured from the line of sight, or what might be called the 'axis' of sight.
As the angle changes, even though the light intensity stays the same the output of the cell puts out less and less as the angle
increases away from perfectly perpendicular. This of course provides a means to measure the differential angle away from 'perfect'.
Not only that, but using three such cells (ideally matched or calibrated) perhaps angled slightly away from each other will provide
a means of measuring the angle of a point source of light anywhere in the semi sphere which is basically the sky.
For example, with only two cells instead of three which are lined up west to east and tilted slightly toward the west and east
as the sun rises in the morning the west directed cell would sense a much higher light level than the cell pointed east and that
would tell the system immediately that the brighter spot was due west. Add another cell off the west/east axis (maybe point it
south) and the system has a way to detect the change in solar pattern over the entire year. That's three sensors, not bad.

The only catch here is that i dont have any hard data to say just how well this works in practice. Theoretically this kind of
system would be able to track a point source with pretty decent accuracy, much better than required to get a 10 degree
accuracy in the setting of the direction of the whole panel, but there is a chance that the atmospheric conditions could
diffuse the light and mess up the readings, unless of course that that scattering still affected the sensors the same
way it would affect the panel (which is most likely). This means it would have to be tested.

Here's a quick example...

Say we point one sensor 80 degrees from the horizontal toward the west, and the other sensor 80 degrees from the horizontal
due east. When the sun rises in the morning and clears the horizon, lets say it's at 10 degrees. That puts its angle to the
west cell at 70 degrees relative to that cell. Since cos(70) is about 0.34, that means the west cell would put out 34 percent
of it's maximum output. Now since the east cell is 20 degrees from the west cell, the angle from the sun to the east
cell is 70+20=90 degrees relative, so the output from that cell would be near zero. This would of course mean that
the panel should be tilted toward the maximum west.
Now as the sun moves to 20 degrees, the west cell incident angle is 60 degrees and the east cells is 80 degrees,
meaning west puts out 0.5 of max and the east cell puts out 0.17 of max. These two measurements are then
run through an inverse cos algo and the angle of the sun is calculated.
Alternately, only the cell whos output is highest is used for the calculation, so in the case of 0.5 of max the
angle would be cosinv(0.5)=60 and subtract that from 80 and we see that the sun would be at 20 degrees west.
Using the highest reading only might work even better because it would reduce the scattering effect.

[jwax]

solar trakin wrote,

well commercial trackers use both the chronological(suns supposed to be position) method and sensor based method(highest intensity in the sky)

Do you have any links to commercial trackers that do this?
Thanks!

[SolarPulse]

http://www.solazone.com.au/tracker.htm

http://www.energymatters.com.au/lorentz ... th=148_167

http://www.energymatters.com.au/lorentz ... th=148_167 -->these guys use optical tracker

i used to have links of couple of other optical sensor based tracker manufacturers, will post em as soon as i find.
cheers.

[MrAl]

Hi again,


This statement is interesting by one of them...

START QUOTE
This is a two axis tracker, that does not use light sensors, but instead is based on a timer.
The tracker does 15 moves each day, and even in very cloudy days, the tracker will be facing the right direction, ready for when the sun shines through.
END QUOTE

[smachin]

Bear in mind that you can determine the position of the sun if you know your exact time and location using many well known
algorithms, e.g. the solar positioning algorithm : http://rredc.nrel.gov/solar/codesandalgorithms/spa/

Once you know the sun's zenith you can work out an angle to drive the panels to. This is the technique used by many commercial
large scale solar trackers.

Using this type of system as opposed to actual irradiance sensors may give you more accurate tracking and you won't have to worry
about keeping the sensors clean etc.

Feel free to email me for any more information.
Sean