Photovoltaic cells 1

[Mtronics]

Hey everyone, my boss has recently purchased a PV cell our plant, one day while speaking to him about it i mentioned that these cells could be maximised to ensure that they are working at their peak. he said that sounded like a great idea and he would love me to look into it. the problem is while i know theres programs you can run to do this, i really dont know where to begin. i feel if i could do this, or come up with a solution i could be in line for a promotion. any ideas?? we have a software package available called 'LabVIEW'.

 

[OperaHouse]

JUST ONE?

And maybe he will just give you a simulated promotion.

 

[Skogsgurra]

Forget about any program. Learn about PV physics first. Labview (which by the way is very capable, wide-spread and not just "a software package you have available") surely can handle any measurement (with suitable transducers and interfaces), simulation and real-time control you will ever need to implement in a PV plant.

Most (all) PV panels have a set of properties that you need to understand. The current-source property (current more or less independent of counter-EMF) and the temperature dependent voltage (higher voltage at lower temperatures - old diode equation) are two fundamental properties. Then, to maximise output, learn how to operate at the knee-point. Manually and automatically.

When you are this far, you can start doing your simulations. But if you get caught in an unknown (to you) SW package before you know what you need to do, then I'm afraid you will not get anywhere.

I also highly doubt that your reason for this exercise should be to please your boss. At least, you shouldn't be so open about it. BTW, Matlab and Simulink and quite a few free simulation programs are also very capable.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[VEBill]

Let me try...

Do you mean optimally aiming it?

 

[Skogsgurra]

I think it is about MPPT (Maximum Power Point Tracking), which usually means that one tries to operate the panel at the knee in the U/I curve.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[OperaHouse]

It is worth mentioning that the power point really doesn't change much except for temperature. In small stand alone systems an advantage can be obtained by using a switching regulator that monitors the input voltage and prevents the voltage dropping below the nominal power point of the panel while limiting the maximum voltage. This can be added at next to no additional cost to a project and likely savings from a smaller panel. Not everything that can be calculated should be.

 

[Mtronics]

Ok ive been doing some research, can I control the maximum peak power solely using labview? Or do I need to create a boost circuit? And dc-dc converter? Dc-ac converter? If I purchased a RIO card for interfacing with labview could I write the algorithm on labview and control from there? I could possibly implement a sun tracker. This seems more simple, by using a sun dial and photocells I could track sun and drive a stepper motor perhaps. But the main aim is implementing peak power tracker.

 

[VEBill]

You'll need a second set of solar panels to power the control systems on the first. ;-)

 

[OperaHouse]

There are off the shelf products that can do all the functions you need. I can't imagine a practical use for labview in solar control other than proof of concept. Your thoughts are hampered since this appears to be the only thing you know. I suggest you take some time researching what is happening in the solar field. Then you could ask something better than a shotgun question.

 

[IRstuff]

I think you misunderstand what Labview does. It's a program designed for graphically controlling instruments and sensors. It has no PV ability beyond what instruments you attach to it and what you program it to do. You can think about Labview as analogous to Visual Basic for Applications (VBA) for Excel or Word.

So, yes, if you add some sort of sun sensor and some sort of servo control, you could conceivably write a routine to track the sun, but you'd need a two-axis pan and tilt since you need to adjust both the azimuth as well as the elevation.

However, you will be re-inventing the wheel, so to speak. There are commercial telescopes that can track stars, and the sun is just a star. I would look for a commercial system designed for PV panels. Note that your pan and tilt needs to be able to actually hold the PV panel, so a particularly large PV panel would not be movable. Note also that star trackers as described don't actually track the star, since the motion of the earth and the stars are predictable to a high level of precision. Ditto, the sun, i.e., there are programs that can predict the position of the sun, given the date and time. There's actually no need to implement an actual sun tracker.

TTFN
faq731-376

 

[eeprom]

Well, I've done quite a bit of this, so maybe I can be of help. There are two methods of maximizing the power output. The first is by using a solar tracker. These will be dual or single axis. The single axis tracks the azimuth angle of the sun (azimuth means the daily change measured horizontally). The dual axis also tracks the sun's declination (vertical). A single axis tracker will improve performance by about 32%. A dual axis will improve performance by as much as 40%, but the closer you are to the equator the less valuable the second axis is. You can make a single axis tracker easily with a stepper motor and a few photo diodes.

Now the second aspect of power maximization is to monitor the PV's operating point. As has been mentioned, the PV has two important parameters: Vmp, and Imp. These are voltage at max power and current at max power. Ideally you want the PV panel to operate at those values, so you will have to alter your load to match those points. It gets complicated because those values only matter with full sunlight. If you have a hazy day, your power goes way down, and you will have to adjust your current output to maximize what the panel can do at that input power. This is complicated. If your panel is grid tied, you can purchase a micro-inverter. This is an automatic synchronizing device which does the power maximization for you.

Another item that would be beneficial to your cause is a pyranometer. This is a device that monitors the amount of solar radiation which is available at your location. You can use this to determine what your input power is. This device is about $250, made by Li-Cor.

A good resource for PV power calculations is PV-Watts. This is a site made by NREL. You can use this to determine (ideally) how many kWh you should be able to generate at any given location.

Good luck,
EE

 

[LionelHutz]

How do you know they aren't already optimized?

There are all kinds of PV grid-tied inverters that do MPPT and every direction I look around here I see 5kW and 10kW PV arrays on solar tracking stands.

You can make sure the PV panels are 90 degrees to the sun at solar noon if you don't track the sun with them. This usually requires changing the tilt a few times a year to at least get close.

 

[VEBill]

There's another option: Take the budget for all these complications, and use the money to buy more (dumb, fixed, reliable, essentially maintenance-free) solar panels.

 

[eeprom]

On a practical level, I agree with VE1BLL. But with a caveat. Do you want to do research, or do you want to minimize effort and cost? These are two different goals. Research is about learning; practical application is about costs.

The cost of a solar tracker is about $3 to $5 per controlled watt. You can buy a panel for less than this. Trackers are fun, but they are not cost effective. I'm an engineer. I like things like solar trackers. But PV panels are very cheap. However, it should be noted that if the goal is being practical, solar panels are not cost effective either.