IAN LANG ELECTRONICS
In this section we look at an increasingly popular subject, that of solar power installation systems. First of all we focus on the smaller systems, and then go on to more complex topics such as concentrated beam solutions.
In the United Kingdom solar power is a minor technology and likely to remain so as the bare fact of the matter is that we don't get a great deal of Sunshine, but in other parts of the world the technology is becoming increasingly important. Spain and China have large scale power stations, as does California, and so the technology does have long term implications for the future of electrical and electronics too, as more and more portable devices are becoming available that either work directly or are charged by light.
The topmost branch of the tree is of course the photovoltaic cell or groups thereof, or the solar panel as such are often called. Let's begin there.
A solar panel is in fact a number of photovoltaic cells arranged in such a manner as to give a specified output of both current and voltage, and hence power by multiplying the two together. So, the first thing we need to look at is how a cell works. First, the simple explanation:
Photons from the light source (in our case sunlight) are absorbed by semiconducting materials,causing electrons to be knocked loose from their atoms which can move in one direction only.
Yep, that's right- it's actually exactly the same as a battery and follows all the rules that batteries do.
Now strap yourselves in because here's where Clever Clive and the Sciency Misfits take over because we are going into the field of solid-state physics (not as bad as quantum but nearly as weird):
As you know if you have read earlier chapters on this website, atoms have electrons. These electrons are said to orbit the nucleus in energy levels. The levels in a good conductor such as a metal are full up to the point of the valence band, which in a good conductor will have few electrons, and the trick is to knock electrons out of the valence band and into a conduction band (the next level, where they move) and thus create a hole in the valence band into which another electron can move and get a chain that way.
In a semiconductor the process is slightly but not much different, in that you have to excite through a band gap between the valence and conductance bands, and this is the job of the photon in a solar cell. Consider the diagram below:
Free Electron Energy
Once a photon strikes, the energy in that photon excites an electron up into the valence band. A band gap exists between the valence and conduction bands and the bigger this gap is the better the material is at insulating. If the energy contained in the photon is sufficient, the electron can cross the band gap into the conduction band, where it is free to move to the next atom across. The next atom across accepts it into the valence band because it's just lost an electron too, and the whole shebang starts again.
If you want to at this point you can read about things like the Boltzmann constant and the Pauli Principle but don't blame me if you do and your brain tries to escape out of your ears. This is strictly for the people who like to sit in laboratories and seldom see a screwdriver and for whom a spanner is uncharted territory.
Silicon atoms have fourteen electrons in three levels of energy. The first normally has two electrons, the second eight and the last only four. If we used just silicon, it wouldn't conduct, because there's nowhere for a hole to exist in the first place as each atom has a steady covalent bond with each other. So we have two lots of silicon in a solar cell, each doped with impurities. The first is doped with phosphorous (nasty stuff this, don't eat it) which has five electrons in the outer shell, and the second is doped with boron, which only has three. Phosphorous gives us the negative side, and boron the positive, because there's plenty of charge carriers going spare in that phosphorous.
When you put them together, you get a PN junction and all the electrons on the N side can cross to the P side. This creates a current, but not for long, because eventually an equlibrium is reached and no more electrons can cross. Until a photon with sufficient energy comes along. That really puts the cat amongst the pigeons. What happens is that the electrons on the P side are excited and move, leaving gaps so more electrons on the N side can shift along. So they do. This causes a current, and between the N and P sides there's a capacitive electric field and this causes a voltage. So now you've got a current and a voltage- multiply them up and you get power.
BUT- the problem is that silicon is really, really shiny. You could if you tried make a mirror out of it, and that's no good because it bounces all the photons away. So, on top of the active layer you have to put an anti-reflective coating and that's why they all look blue. If it rains or sand gets on the coating, it'll wear off, and so now you have to protect that too, and so you put a glass cover on to stop it happening and encase the whole lot in a nice frame. Voila.
Just to confuse the issue the silicon substrate can be made in different ways. Commercially, the three you are most likely to see are:
1. Monocrystalline which is expensive but does in fact convert the most energy, on average about 17 % of photon energy is converted into useful electrical power. Monocrystalline cells are made from silicon having a uniform crystalline structure with no grain boundaries in exactly the same way that chip substrates are made.
2. Polycrystalline, which is made by heating ingots of non-uniform crystal and are cheaper by far, but unfortunately only an average of about 10 % is converted into useful electricity.
3. Amorphous or thin film, which is incredibly cheap but truly awful. Vagaries in manufacturing mean that no real meaningful data can be gathered about performance. Some manufacturers claim 10 % but it's unlikely you'll get anywhere near it. It's used for things that have a very low power requirement, such as calculators. If you see this on the roof of a house, ask the owners which firm did the installation. That way you can avoid them.
Looking through those figures, you'll find the conversion rate is not very good. Even the very best panels made to industrial specs and using the latest techniques have achieved just over 40 % and they are too expensive to be commercially viable. The best monocrystallines can achieve about 25% and they are far too expensive to be widely commercially available.
The trouble with light is that it isn't all one wavelength and so the photons don't all have the same energy.
(And no I don't know how light can be a wave and a particle at the same time either. All the Sciency Misfits say it is and there's no point arguing with them since they just babble theoretical physics at you until you submit. Sometimes they use the word "quantum" and then it's all downhill because if you keep arguing after that they break out complex mathematics. At that point you either jump out of the window or spend the next five hours in utter bewilderment and the next two weeks with an odd feeling that all the atoms are going to pounce on you as soon as there's no witnesses.)
I digress. Since not all photons have the same energy, some will provide too little energy to break the bonds and move the electrons. Some will actually have too much and will pass through. The ones that have twice as much energy will actually cause two electrons to shift for the price of one, which would be a mighty bonus except that aren't enough of them. The right energy is about 1.4 eV (electron-Volts) for silicon and it's to do with that band gap. You could change it, but then you'd reduce the voltage output and so there's no benefit. There's also resistance in the cell itself causing more losses.
In fact a cell can put out about 0.45-0.6 V and a small current, but the more current drawn the lesser the voltage (rather like an unregulated transformer) and as it gets hotter it gets less efficient, which is something of an embuggerance when the whole point of the thing is to point it straight at the Sun.
In a panel, the cells are arranged in arrays. Firstly, putting them in series gives more voltage (you add up the voltages taking the lowest output as your base) and putting them in parallel gives more current. If you now parallel several up, and series connect that array to a similar array, you get an increase of both voltage and current. You also get an increase in power, and an increase in heat, leading to a bit more loss. C'est la vie.
The most important thing is the density of light hitting the panels, not the source. Small panels work just as well in artificial light from an incandescent bulb as they do in the Sun. As the sun moves across the sky, it comes more directly in line with your panels in both the horizontal and vertical components, and you'll find that you get most efficiency out of your panels when the Sun is directly on them. This has several meanings for us in fitting them. Firstly, in the Northern hemisphere, the Sun goes round from East to West via the South. In the Southern Hemisphere, it does it via the North. I can't imagine what it does on the Equator- does it just sit there and look a bit lost?
I digress. In the United Kingdom this means at Noon (or thereabouts) the Sun will be at its highest point in the sky and southerly. In Australia at Noon it will be northerly. So, if we're in Melbourne, we need our panels on the roof facing north. If we're in Godalming, we need them facing south. (If we're in Manchester I don't think it matters since every time I've been there it's been raining cats and dogs and the sky has been dark...)
This of course means that although your solar panels will work on the light of day when the Sun has risen and when it's setting, the greatest efficiency will be only for a few hours. In high summer, you'll get a decent output from about 6 in the morning to maybe 8 at night. In midwinter from ~11 until ~2. Unless it's cloudy. A good layer of cloud reduces the density of light so much that your panels will work at well under half their efficiency. In fact, I've seen a small panel drop from a 12V output with enough juice to power a small water pump to less than 2V with a current in milliamps, just because a bank of cloud has covered the Sun. The water pump of course stopped immediately. When the cloud passed over two minutes later, the pump started again.
The majority of panels are mounted on the roof. This means that your roof has to have the right inclination and general flatness of profile to accomodate the panels as well as being in the right direction. A normal, south facing 45 degree slope is good, a hip or mansard roof is not, and whilst you could put them on a flat roof it may not be worth the cost. Your roof must also be able to bear the weight- they're silicon, glass and steel in quite big quantitities and that's a lot to be slapping on your trusses.
Now, what we've got to remember is that the panel is only the beginning of the system, the batteries if you like. All it (or they) do is provide a voltage and a current, and moreover it's a direct current (DC). If you want to be able to run your entire household on solar power, you'll find that most of your heavy appliances rely (in the UK) on 220V AC as supplied by the National Grid.
Your panels put out a few Volts DC. It doesn't match. So what we now need is a system that makes it do so. Here's one:
Battery DC Load
The above diagram is a complete stand-alone solar installation and not all installations will need all seven components. Some of them may be familiar, others may need a bit of explanation. Over the page, we'll start with direct loads, get them out of the way and go on to the more complex pieces.