Domestic Solar Panels

-are they worth the outlay?

Quite a few people ask for advice on rooftop solar panels. Meaning the electrical or photovoltaic variety, not the type that provide hot water. Usually they've had a door-to-door salesman turn up and try the hard-sell, promising them all kinds of advantages all the way up to that holy grail, free electricity for ever.

Now, I don't have a solar installation so it might be worth asking someone who does for some firsthand figures. However, I have used small-scale solar panels for battery charging and the like. What I can tell you is that whilst a panel will deliver a reasonable current in bright sunshine, in the typically cloudy British climate the average output is not all that much.

Disregard the sales people who will tell you glibly that, 'Modern panels still give output output in cloudy conditions.' Of course they give an output. Question is, how much? Are we talking kilowatts, or just a few watts? Test with a 20W monocrystalline panel show that output drops to a fifth or a tenth of rated value in overcast conditions. This is unsurprising in view of the difference in light levels. 

The key factor in solar performance is insolation. Insolation is the amount of raw power striking the ground, from sunlight, at any given time. As the sellers will be keen to tell you, the sun provides us with nearly a kilowatt of energy per square metre per ground, and all for free! So what are you waiting for with that amount of energy available? Sign up now!!

What they probably won't mention is that (a) the kilowatt figure is at the equator where the sun is (nearly) directly overhead at noon, most of the year. In the UK's northern latitude you might be looking at nearly a kilowatt in summer, but a LOT less in winter when the sun is at a low angle. Also, (b) that in cloudy conditions the amount of sunlight which actually reaches the ground could be less than a tenth of that. Suddenly we start to see a rather more pessimistic projection of actual yields.

Fortunately, NASA do provide some fairly detailed figures for UK insolation:

The amount of raw energy a square metre of panel will intercept in an average day.
The electrical output will be about 16% of this.

Let's do the maths using these, and see what answers come out:

For a square metre of ground (or roof) NASA give typical insolation figures of 4.5 kilowatt-hours ('units' in supplier-talk) per day, in British summer. That is, allowing for weather variations we would expect our square metre of ground to soak-up the equivalent of one kilowatt of heat and light for four and a half hours, or (more likely) half a kilowatt for nine hours. However, the insolation figures for winter are very much lower. Exactly how much lower depends how far North you are, but you can expect less than half a kilowatt-hour per typical midwinter day. If we average the monthly yields, we find that the mean insolation is around 2.42 units per day, or 883 units per year.

A typical roof installation might be twenty square metres, as a five metres by four array of panels. A square metre of panel being rated at about 150W output, such an array would often be described as a 'three kilowatt' installation by the seller, which refers to the maximum possible output in very bright sun.

This array will soak-up 883 x 20 or 17660 units of raw solar energy over a typical year . However, the efficiency of solar panels at converting light to electricity averages around 16%, with the best available managing about 20%. So, assuming 16% you would be looking at 2825 units of actual electricity from the panels per year. Here we are also assuming an ideal placement of the panels to intercept the sunlight. Real situations may fall short of the ideal; by how much is quite complex to estimate.

There is a moral issue here, in that solar subsidies are paid for by other electricity consumers. Thus, for each kilowatt-hour you generate, 15p is actually wealth which you have created by way of saving energy, but the rest, more than half, is effectively charged to your neighbours' electricity accounts.. Without their consent!

This isn't the end of the supply chain, because the power from the panels has to be converted from around 40v DC to 240v AC for domestic use, and there will be some further losses in this stage of processing. The conversion efficiency varies depending on several factors, but for argument's sake let's assume 80%. So, your actual usable yield would be 2260 units.

Another less accurate but useful way of obtaining a ballpark estimate for output is to assume that in the UK, panels will receive the equivalent of two hours of good sunshine per day when averaged over a year. So by this yardstick our 3kW array would receive 6kWh (6 units) a day, or 2190 units a year. You will notice this is similar to the figure obtained from theory, including conversion losses. Of course, both figures are subject to a considerable amount of variation, such is the nature of UK meteorology!

So, what's it worth? At standard retail tariff, a unit of electricity is worth around 13p to 15p. Thus, your 20sqm array should return about £339 actual worth of electricity per year, assuming 15p per unit. .

That is, of course, at standard rates. For approved installations you may receive a subsidy known as the feed-in tariff. This provides you with an additional payment for every kilowatt-hour your solar panels generate, regardless of whether you use it in the house, or export it to the Grid.  If you export all or some of your solar energy to the Grid, you also receive a smaller 'export tariff' payment in addition to the basic feed-in tariff.

When solar subsidies first started, feed-in tariffs were set at about 45p per unit. Since then they have been progressively reduced to 14.9p per unit, as of mid 2013 on. The export tariff remains at 4.64p per unit.

The average value of the electricity you get from a domestic solar installation, per day.
Subsidies will (typically) somewhat more than double these figures.

In mid-2015, solar panel costs have now dropped to the point where they are well under a pound a watt, in fact some 250W panels are available on Ebay for just over 60p a watt. Thus the economics of the actual hardware are a lot more favourable than they used to be a few years back. The other side of this is that feed-in tariffs are a lot less than they used to be.

As regards the economics, it is not hard to see that at basic rates without interest it would take nearly eighteen years to pay off a £6000 installation cost, twelve if you can get the cost down to £4000. This is hardly economic. However, with the maximum subsidy and a bit of luck with sunny weather, that drops to just under six years. Maybe take eight as a realistic goal with full subsidy. Still a long-term investment, but at least more realistic. This page explains the FIT rates quite well.

A calculator is available on the Energy Saving Trust site, which seems to give reasonably accurate results.

After payoff, the return on your investment will depend very much on whether subsidies continue to be paid, or not. The government has guaranteed the feed-in payments until 2020 for existing owners. New owners may of course be subject to different conditions, so check the small print. With the full subsidy you would earn up to £1000 a year depending on when you signed-up. Without, £339 a year. This assumes current electricity prices, of course. If prices rise you would benefit more, but if they come down you will benefit less, or may even lose your capital.

Considering outside influences on your investment, a national switchover to windpower or other costly renewable sources would be likely to create a massive rise in domestic electricity prices, plus widespread shortages and blackouts, in which case you would benefit even more from the virtually free power from your paid-off solar panels. But, if shale gas or (say) a nuclear innovation provides an abundance of cheap power, you will lose out. Right now, no-one can answer that one.

As a ballpark estimate for UK conditions, you should get around twice the installation's rating in kilowatts, as electrical output in kilowatt-hours, per day, averaged through the year. So, a 1kW installation should give around 2x365 or 730kWh per year.

There is also a moral issue here, in that solar subsidies are paid for by other electricity consumers. Thus, for each kilowatt-hour you generate, 15p is actually wealth which you have created by way of saving energy, but the rest, more than half, is effectively nicked from your neighbours! Is this socially acceptable? Good question.

The other oft-overlooked issue to consider is that solar tech is in its infancy, and it is quite possible that in ten years' time (or even five) cheaper and more efficient solar panels will become available. If you are one of the early adopters you will then be faced with a dilemma: Do I keep my old-model panels until they've paid for themselves, or do I replace them with the more efficient designs? Keeping the old ones means accepting a lower output, but replacing them means waving bye-bye to part of the planned return on your original investment. Thus there are arguments for waiting until the technology is more mature, and costs per installed kilowatt are unlikely to reduce much further.

This reminds me of the SF poser raised by Arthur C Clarke: If it becomes possible to build a starship that could reach Alpha Centauri (or wherever you think might be worth visiting) in ten years of travelling, do you get together a crew and set off rightaway? Or, do you wait for better technology that might take a few years to develop, but could perhaps get you there in half the time? The issue here is that it would be highly embarrassing, to say the least, to be passed enroute by a much faster ship, launched a few years after yours.  In fact it would make your whole mission pointless,  since the later but faster ship would not only get to Alpha Centauri first, but would  then be able to return to Earth many years ahead of yours. That's assuming you didn't just eat humble pie and hitch a lift back on the newer ship, of course.  The point this anecdote tries to make is that premature deployment of any technology can be a costly blunder, especially if there is still plenty room for improvement.

I hope this gives some idea of the financial aspects of domestic solar electricity. I don't guarantee the accuracy of any figures given here, but they are based on established sources and should in principle be correct.

For more information and a second opinion, see this JRC solar calculator. The winter figures seem a little optimistic by NASA's reckoning, but overall the results are similar.

The actual installation cost:

Prices are steadily coming down, and as of writing about £6000 seems typical for a full-size domestic install. Of that, about half will be the solar panels. Additional costs are mounting hardware, connectors, cabling, inverter and metering. Plus of course labour.

-Which brings us to the inevitable question for tech guys like us: Can I DIY my solar installation? In Scotland yes you can, at least in principle. In England you are liable to get caught-up in the red tape created by the 'Part P' building regulations. Materials are available on Ebay, and solar panels currently cost around a pound a watt of full-sun output. However, a DIY install would have to make do with the actual value of the electricity produced, since the FIT is only paid to installations done by accredited installers. So, a DIY install would have to make do with less than half the revenue, but would of course save on labour costs. And, at least that would be revenue honestly earned, and not surcharged to the rest of the UK's electricity bills. Then again, you would need to buy specialist tools to do the install, and using them for only one job might not be too economic. Plus of course you don't want to create an electrical hazard or roof leak -or worse a roof collapse through underestimating loading, since the panels are surprisingly heavy. So, unless you are a reasonably skilled DIY'er it's probably best left to the pros.