Anything that moves will one day break. I hesitate to move away from the reliability of a static panel. Adjusting the drip rate in this scheme would also be constant battle of weather, temperature and season. You'd never get it perfect.

Compared to a simple solar setup it may be more efficient, but compared to more complex, yet still static, combinations of a panels and a reflectors I don't think the improvements are worth the extra complexity of moving parts. I'm thinking of rigs like this: http://www.builditsolar.com/Experimental/Reflectors/Reflecto...

I've had a rooftop solar system on my home for about ten years ago. I would have never considered an active tracking system, as the wind would have wrecked it long ago. The only maintenance I have to do is, at most, to hose the panels off about once a year.

If you're trying to build out solar in a location where space is at a premium, you're doing it wrong.

This system doesn't seem very expensive or difficult to repair. And in any case, the worst-case scenario is that the solar panel becomes static.

The worse case scenario is that the panel breaks. Is the framework actually designed to handle the weight of the water?

I would presume so, as that is what it has been designed to do.

I worked about a year in a project to design a solar tracker and we pretty much came to the same conclusion.

We hit several obstacles but in my opinion the two main reasons for the failure of the project were:

1. Safety: We underestimated how a lack of regulations left us in a legal limbo

2. Reliability: We underestimated how quickly outages and repair cost eat the profit

I don't say it's impossible, other companies (like Solon for example) sold single and dual axis trackers back then and they are still selling them today. Our design was much simpler than the existing solutions and we still ran into the reliability trap.

Yes, tracking has got to be super-simple. Flowers do it; maybe one day we'll use them somehow. Flower power.

But the efficiency difference is dramatic, especially for monocrystalline. I got full power in the middle of winter near sunet (4pm) by tilting the panel almost vertically.

Consider the effect of solar ray incidence on Earth: it's why the tropics are hot and the poles are cold. It's the difference between summer and winter.

> Yes, tracking has got to be super-simple. Flowers do it; maybe one day we'll use them somehow. Flower power.

Flowers are not simple. They're insanely complex devices built entirely by and from nanomachines. We're barely beginning to understand what's going on. For instance, we found solar tracking to be a set of clever photochemical reactions that respond to particular wavelengths and disrupt the natural growth of the plant in such a way that it follows the sun. Mind you, that we know about those reactions does not mean that we actually understand them in perfect detail.

Nature only looks simple. In fact, the levels of its complexity are so great that for the most of mankind's history we didn't even suspect what's going on.

The efficiency benefit is great, but the extra cost of a rotating frame in 2015, after years of panel price reductions, may end up being more money than just adding extra panels.

Cool link! But would you add a motor and microprocessor to assist with the monthly angle adjustment, please? :p

Seriously, this is a new idea to me. It looks like the chief weakness is it doesn't really help in the winter, which is exactly when you need the most efficiency.

Yes, you must size your panels for the worst case, not the best case.

When it's overcast, the best position is pointing straight up, eg fixed at the correct angle for that latitude.

To put it another way, a tracker optimises your panel in sunny conditions, which is not worth doing. That's when you have a surplus of charge.

That's a good point. For a given deployment site, it might still be worth optimizing for sunny conditions though - depending on the number of sunny days and their distribution across the seasons, the availability/cost of storage from one sunny winter day to the following overcast day, and whether or not you care at all about peak output.

If you can afford to get more panels than strictly necessary, this solution isn't very good. But I guess for households in poor countries it could really help.

So your rig will generate absolute zero in the middle of the day when the sun is above your head.

This is called a clepsydra (water clock) tracker.

A group at New Mexico Tech added water pumps to the technology to take out the need for manual refilling.

What people tend to forget about trackers are they cost money and have to be massive to hold their angles in the wind.

They are also less space-efficient users of land than fixed mounts, which is why standard PV's never tend to use them. It also costs less per watt to buy a second panel and fixed mount than spend the extra on any kind of tracking mount.

As you said, the price of panels relative to the price of trackers is such that trackers don't make economic or engineering (complexity) sense.

I think we we potentially getting there with energy storage as well for off grid installations. Rather than trickle charge a storage system that can supply short bursts of massive power (think running a welder off of solar), purchase more generating capacity and feed it directly into the inverters and use storage for the minimal case. Excess can be shunted into a opportunistic load.

Another issue with trackers in the wind is the failure condition can be catastrophic; an actuating arm smashed through the middle of its little array is not a pretty sight after a wind storm. Presumably this can be engineered around, though. I think the absence of this category of flaw is a distinct advantage to the gravity-powered design of the OP.

Ah, before I read the article, from the image I thought maybe it was bottles on both sides, and using the sun to evaporate the water in the lower bottle until it reaches the higher bottle, causing it to tip. Thought that was pretty clever! But probably not accurate. Would be beautiful to design a closed system that works well like that however.

Actually there is a simple mechanical gear that can translate constant speed rotation into a perfect two axis sun tracker. If you'd combine this with the contraption from the article you'd get a gravity-powered two axis tracker.

A human has to refill the bottles.

Another mechanical approach that is relatively cheap and doesn't require refilling bottles is to use a closed system of a hydraulic cylinder and two exposed freon tanks, one facing east and one facing west: http://www.motherearthnews.com/diy/solar-tracker-zmaz77ndzgo...

These mechanical approaches are generally awesome because they hold up well in bad weather.

If you wanted to go all out I don't see why you couldn't combine the gravity system and the freon system and track on 2 axes! (Anybody got a clever idea how to make one system or the other neatly handle the axis corresponding to the height of the sun?)

What about an array of photodiodes like this:

  o o o
  o X o
  o o o

This is in an enclosed box with a hole over the center. Here, "o" represents a photodiode and "X" represents the center of the hole. When the solar panel is pointed directly at the sun, the sun is at position X and no photodiode is lit up. As the sun changes position, eventually it will light up one of the photodiodes. That completes a circuit which uses power from the panel to drive a motor, which is attached to a worm gear and rotates the panel towards the sun. Eventually the sun will be pointed at the X position again which breaks the circuit and stops the rotation. Each diode completes a circuit, driving one or both motors (in the case of the corner diodes) rotating them appropriately.

It'd be interesting as well to have some sort of mechanism which, when the panel receives power, keeps it locked in place. When the panel has no power for a period of time (ie: after sunset) the mechanism disengages and the panel falls back to its default position pointing towards sunrise.

Finding vertical & finding north is not a problem, you can get that trivially in a tiny MEMS IMU if you don't want to do it manually, and from there determine the optimal path for solar tracking without added variables, supposing you have the timezone set by the distributor (or use a GPS for that at a few dollars a chip).

The cost is in the rotating frame and the motors.

Good ideas!

I had in mind a purely mechanical mechanism that takes the linear motion of the hydraulic piston or the water clock and turns it into a parabolic movement. I'm a software guy and could not find such a mechanism easily on Google, though.

In a country like India, especially in the villages, there would mist likely be enough free labor to do something like this.

Free labor? Slavery?

i think analogous from your observation could be: fill water bottles once in the morning or additionally use mind, motor skills, and body, lose the free use of a hand, resources to wave a fan in front of your face all day;

Or, humans use electronics and motors because it actually is more efficient than getting a human to do the work.

"Gravity powered," not "fully automated."

I think this is pretty clever.

"It is 30 times less expensive than conventional motorized solar panel rotators" if you ignore the fact that getting a human to do something is super expensive.