Renewable Domestic Power: Solar Collectors

There are two main families of solar collectors: thermal and photovoltaic. Thermal collectors turn sunlight into heat, and photovoltaic collectors turn sunlight directly into electricity. Thermal collectors are cheap, and photovoltaic collectors are much more expensive, but they also require much less maintenance.

The Sun Moves

The Sun does not stay in one place relative to a solar collector. During the course of the day it travels across the sky, and during the year its track rises and falls relative to the horizon. This affects the amount of energy that can be collected.

The amount of energy that can be collected can be calculated by taking the amount that is collected when the Sun shines directly on the collector, and multiplying it by the cosine of the angle between the Sun and where the Sun would be if it was shining directly. The values of cosine remain near 1 until an angle of maybe 30^o is reached: so if the collector is facing due south, the energy is nearly the same between 10AM and 2PM. By 45^o, cosine has dropped to 70%, and by 60^o it has dropped to 50%. So the energy is more than 70% between 9AM and 3PM, and more than 50% between 8AM and 4PM. After that it rapidly drops to zero.

A similar argument applies to the effect of seasons. On midsummer's day, the Sun is 22^o higher in the sky than at the spring and autumn equinoxes. On midwinter's day, it is 22^o lower. A solar panel aligned to catch the Sun on midsummer's day will only generate 70% of the energy on midwinter's day.

The geometry of house construction and the lattitudes we live at tend to set the limits of how we can align our panels. We want to live at 53N, which means that the Sun will rise an average of 37^o above the horizon at the Equinoxes; 59^o above the horizon on Midsummer's Day; and 15^o above the horizon on Midwinter's Day. The houses we are considering buying have roofs pitched at about 45^o and the walls are (we hope) horizontal.

It can be seen that a wall panel collects most energy in winter, whereas a roof panel collects most energy in summer. If we are planning a solar photovoltaic collector, it is best put on the roof, since this generates most energy in summer, when the hours of sunshine are longest. But if we are also planning a solar thermal panel, to provide hot water and heating, it would be better placed on the walls, where it generates most heat in winter, so that the space on the roof can be left for photovoltaic, and so we can avoid boiling the heat store in summer.

Thermal Collectors

A thermal collector works by shining sunlight onto an absorbent surface. This surface then warms up, until the absorbed energy is equal to the energy lost. Broadly there are flat panel collectors, which work by exposing a large black sheet to sunlight, and focussed collectors, which work by a parabolic mirror with a smaller black surface at the focus.

A Panel Design

A thermal collector normally consists of an absorber, a backing insulator, and a window. Sun shines through the window, is absorbed by the absorber, which gets warm, and water circulates through the absorber to carry away the heat. The absorber also loses heat by infrared radiation, and by conduction. Foil and an air-gap on the backing insulator can all but eliminate radiation from the back, but the window is likely to allow the absorber to re-radiate infrared. The design of the absorber and the window is therefore critical to collecting a good amount of energy.

Two possible window materials are considered: SGG Planiform glass, which is glass coated with a layer that reflects infrared radiation back to the absorber; and Corotherm 16mm dual-layer polycarbonate, which is better at transmitting sunlight than the glass.

(The solar factor figure is not for Corotherm, it is for clear polycarbonate sheet produced by CPI International. In fact the Corotherm figure will be higher, since it has two walls, which will be at different temperatures.)

The Sun produces about 1kW of energy on a surface that is perfectly positioned to catch the sunshine. But the observations about the Sun's movement during the day mean that we might reasonably only average 800W per square metre.

That figure needs to be reduced again by the amount that is transmitted by the window. In the case of the glass, that is 616W per square metre; in the case of the polycarbonate, it is 656W per square metre.

To calculate the effectiveness of the solar collector at various temperatures, the re-radiation of heat must be considered. Various coatings can be used for the absorber, but the best sort of material in the price range is something like Solkote HiSorb, which absorbs maybe 94% of visible light but only transmits 28% of the expected IR energy.

By Planck's Law, the power radiated by a black body is:

P = σAT⁴

where A is the area, T is the absolute temperature (in K) and σ (sigma) is a constant, 5.67051(19) x 10^-8 J m^-2s^-1K^-4.

If the panel is perfectly black, and in a room where everything is at a temperature of 300K (about 27^oC) then the panel will be losing and gaining about 460W per square metre. The temperature will remain at 300K, though, because the emitted and absorbed radiation is the same. But if it's different, the temperature will change.

Here are calculations done for the polycarbonate panels mounted vertically at the target lattitude (53N), and at average daytime temperatures for July(18^oC) and December(8^oC).

Here are the same calculations done for the SGG glass. Note the higher summer absorber temperature.

So what does that mean for our practical collector? Our design is principally aimed at providing heating in winter, and for our house we'd aim to put 2m x 11m of panels on the front wall. The winter conditions (between 1.5 and 2 hours of sunshine a day) will give 10kWh to 12kWh of heat per day, on average, which would make a sizeable contribution to winter heating bills (maybe 20% of the total if we insulate to British building regs). On sunny, calm winter days, the solar panels might produce all the heat required.

For the summer, the panels are exposed to typically 4 hours of sunshine a day, and would provide 8kWh per day, depending on the design. That doesn't need to provide much heating, but does need to provide hot water: it should do that without any problem.

The advantages of low-E glass over polycarbonate are felt most keenly in summer. In winter, the low-E glass provides 15kWh to 20kWh per day, which is 25% of the heating bills, but in summer it provides double the power. The useful part is in the winter, though, since the summer hot water requirement is certainly met by either design.

One other factor that may prove interesting is that the low-E glass allows higher temperatures to be obtained in the heat store. That makes it possible for heat collected on sunny days to be stored more, since the maximum absorber temperature affects the maximum temperature in the heat store. However, given the changeable weather in this part of the world, and the wind generation contribution, it seems unlikely that this will be very significant. I guess we need to see the prices of the glass and the polycarbonate, and do some comparison.

A Reflector Design

Now consider a mirror collecting 1 square metre of sunlight and focussing it on a collector of 0.01 square metres in area. Assuming similar insulation properties to the above panel for the collector, in the same 30^oC conditions described above, when 600W falls on the mirror the radiation gained is 600W from the Sun and 4.79W from the surroundings. But now the loss at 90^oC is only 11W for conduction and radiation: clearly the temperature will rise much higher than this. So this arrangement can generate 579W at 90^oC.

Even in winter, with our 0^oC surroundings, the heat loss at 90^oC is only 12W: so the output is 378W at 90^oC in these conditions.

Clearly the reflector design can be used to generate high grade heat: useful for a heat engine like a Stirling, which will recover only 3% of the energy between 300K and 310K, but 20% of the energy between 300K and 360K. So the panel design will give about 10W of work per square metre in winter sunshine, but the reflector design will give about 76W of work per square metre.

The bad news about a reflector design is that the reflector needs to be steered to point towards the Sun. That makes things considerably more complicated, and so expensive and unreliable.

Photovoltaic Collectors

Photovoltaic collectors generally use silicon to turn light into electricity. The best photovoltaic collectors are about 12% efficient: more typical ones are about 3% efficient.

There are two types in popular use: crystalline and amorphous. Amorphous panels are cheaper, but they don't last as long: typically only a few years. Given how long the payback time is for solar panels, that is a problem.

Which brings me on to the most important characteristic of PV solar panels: cost per watt. The absolutely cheapest solar panels I've found are provided by someone called Bullnet and come in at £1.84 per watt. Most panels are around £3.50 a watt. This means that a system delivering 1500w is going to cost something like £2800. If it has to be replaced every three years, that's even worse.

Sunslates

Atlantis Energy Sunslates are a standard sized slate which has PV cells built in. The idea is that the appearance of the roof still looks somewhat like slate, but it's actually a PV array.

For our system, we want about +/-48v, which would mean 34 units as two chains of 17, each delivering about 4.6A at +/-49v. The roof is 14 rows of about 30 slates, apparently of the same size as the Sunslates, so we should be able to get ten strings on the roof, for 46A peak, or 4.5kW installed. That'll not do much for the dumpload, but it'll provide plenty for the batteries and enough to run intermittent appliances.

Of course that's the springtime direct sunshine at midday scenario, when the roof is seeing about 40kW of sunshine: in practice you'd get maybe 70% of that during a particular day. So in an average May day of 5.77 hours of sunshine, we'd actually get 18kWh, which is still quite respectable. Maybe the dump load would have some use in summer.

In December, when the Sun only shines for 2.1 hours a day and the angle means the roof only sees 87% of the direct radiation, the Sunslates would generate nearer 5.76kWh per day. That's surprisingly high -- more than I'd have expected.

Other Options

One wild and wacky possibility would be to use a parabolic "trough" reflector with an air pipe in the middle, and connect that air pipe between the inlet and exhaust of a scrapyard-obtained turbocharger. By bolting or gluing permanent magnets to the shaft and fitting coils, it would be possible to generate electricity. This is a heat engine using air as a working fluid.

A turbocharger on a 2-litre car running at 6000rpm is pumping 100 litres of air a second. Typically it'd boost about 6psi at this speed, which is equivalent to 40,000Nm^-2: about 0.4 of an atmosphere. The approximate energy required for that is 40,000Nm^-2 x 0.1m³s^-1, or about 4kW. So there's plenty of room for some work to be extracted.

If we aimed to get 30% of the work out of the shaft - a good efficiency for a heat engine - then to get 1kW of energy out of the shaft we'd need to collect 3kW of heat. In summer that'd be 5m², and in winter about 7.5m². It would also need a silencer on the output, since it would be noisy.

One way to silence it, of course, would be to exhaust the turbo output into the heat store. This would have the effect of absorbing sound in long pipes surrounded by water; and also the effect of heating the water. As well as getting that 1kW of electricity, we'd also get an additional 2kW of heat.

The only other details would be an oil pump to keep the bearings cool and lubricated, and a bit of electronics to turn the generator into a motor when the collector is hot but the turbine is not running. But even with all these things in mind it is remarkably cheap!

Site Considerations

Here is some meteorological data for London (where I live) broken down by month:

London wind met data is source The Washington Post, and sunshine data from Roehampton.

Here is some meteorological data for Galway (where I'd like to live) broken down by month:

The peak wind speed is gusts of 93kts and 10min averages of 60kts in September. Galway met data is source The Irish Met Office, data for Shannon Airport.

1kt is also 1.86km/hr or 0.517m/s.

These figures were put into a spreadsheet to create estimates of seasonal variation of power. Then we aimed to get at least 125% of our needs in every month, adjusting the powers of the two accordingly. For London the best mix is about 150% peak by solar and 150% peak by wind: for Galway it is about 50% by solar and 200% by wind. This is because Galway is windier and London sunnier. These figures are good news, because a given property in Galway costing the same as in London is likely to have more space for windmills; and more need of heat during the winter, when the wind energy is higher.

The idea is that there are enough solar panels to generate 50% electricity in May, the sunniest month. That means between 10kWh and 15kWh per day, or between 1.75kW and 2.6kW in full sunlight.

This page is some notes on Domestic Power from Renewable Sources, and is written and maintained by Simon. At this stage these pages are constantly under revision. Thoughts and comments are welcome.