Solar Scotland

I live in Aberdeen Scotland where the Sun seldom shines, especially in winter when it is cold and electricity demand is highest. And yet the British Government, supported by the Scottish Government, the European Parliament and by local Government, are subsidising the deployment of solar photovoltaic panels everywhere [1]. These are as likely to be located on North facing as South facing roofs in Aberdeen and I’ve been told it makes no difference. I always believed that for optimal performance of solar PV, the panels had to be carefully orientated towards the Sun. What I see going on around me seems to be a criminal squandering of resources.

It may be hard to believe, but sunshine is going to transform Glasgow into one of the world’s greenest cities [1]

This post was originally part of The Efficiency of Solar Photovoltaics [2] that was posted last week. My core motivation is to examine the wisdom of deploying solar photovoltaics (PV) in a land where the Sun seldom shines. Is this strategy really Green or is it like most other strands of energy policy, simply window dressing?

Figure 1 This solar PV array is mounted on a roof facing due east on a relatively new building in Aberdeen. The picture was taken at 16:00 hrs on 2nd May 2014 on a rare sunny day 😉 Due to the architecture, half of the array is already in the shade and the remainder is so oblique to the Sun it will be capturing barely any direct sunlight. To the left of the picture is a large roof section facing due south. It is inexplicable that any properly educated engineer could construct a system in this way.

The efficiency of Scottish Solar

In my earlier post there was considerable doubt and debate about the veracity of the BP statistics [3] for installed capacity and solar energy delivered. But in light of the discussion in comments it seems reasonable to conclude that at sunny low latitudes – New Mexico, Mexico and Portugal – that the load factor for solar PV is around 20%. That means the aggregate output is one-fifth of the nameplate capacity. At higher more cloudy latitudes where the sunlight needs to pass obliquely through the atmosphere the load factor may fall below 9% in a country like the UK. This is due to reflection of sunlight off the top of clouds and scattering of light through the “deeper” and moister air column.

Sunshine and warmth is not evenly spread across the UK. All too often we in Aberdeen need to listen to the MET office forecasting 25˚C and sunshine in London with 15˚C and rain up north. This perception is born out by climate data that shows the Shetland Islands (way up north) receive about 1100 hours sunshine per year whilst sunny Eastbourne on the South coast of England receives 1900 hours – that is 64% more sunshine than Shetland (Figure 2). The Scottish stations cluster around 1300 hours while the English stations around 1600 and so I estimate that the Scottish load factor is likely nearer to 1300/1600*9% = 7.3%. But since the panels are not even pointed at the Sun, it is likely to be much less. 1300 hours means it is sunny for 30% of daylight time and that seems to be about right.

Figure 2 Sunshine, 5y running averages for 23 UK climate stations [4, 5].

Orientation of solar panels

Does it matter if solar panels face north or south in Scotland? I’ve been told it makes no difference since the panels only require there to be daylight and the installation engineers clearly seem to believe it makes no difference since Scottish solar faces every direction, including south.

Figure 3 All Earth Renewables are a US company offering solar gantries that track the Sun, optimising the efficiency of capturing light.

Figure 3 shows how the professionals do it. First of all choose a sunny locality. And then mount the solar panels on a gantry that tracks the Sun. Not only do you optimise for latitude but you follow the Sun each day as it tracks across the sky. All Earth Renewables claim that the system delivers 45% more electricity than static panels.

In this comment, Professor of electrical engineering at Cal Tech, Dave Rutledge said this:

If I take a 100-W solar panel here in Los Angeles and aim it properly at the sun, I get 60W output. However, when I point the panel end on to the sun, I still get 15W. That is, a lot of light is scattered at our latitude.

Sunny Los Angeles is 34˚N, cloudy Aberdeen is 57˚N. Given that it is only sunny for about 30% of daylight time in Scotland, solar panels are likely working on diffuse scattered light for most of the time, hence it probably really does make little difference which way they face. This is no miracle. It means that the performance is so bad that mounting solar panels on Scottish roofs is only marginally better than leaving the panels in their boxes – and perhaps not even so.

The Energy Return on Energy Invested of Scottish Solar

In my earlier post [2], Pedro Antonio Preto claimed in this comment that a comprehensive study of Spanish photovoltaics conducted by himself and Professor Charles Hall gave an ERoEI of 2.4 : 1. In other words, during the lifecycle of the panel it will produce 2.4 times the energy used in its manufacture. This figure is considerably lower than other estimates reviewed by Professor Charles Hall in an Oil Drum post [6] which fall mainly in the range 4 to 10 : 1, stretching up to 20 : 1 (Table 1). In doing this kind of analysis much depends on the life span assumed for the device, the load factor of its location, performance degradation with time and the boundaries of the energy analysis – whether or not the energy used to mine and transport raw materials, to build the factory and to feed and shelter the workers is counted or not.

Table 1 Review of Solar PV ERoEI provided by Hall [6]. See ref 6 for data sources.

The low number reported by Preto is supported by a dynamic analysis of off grid solar PV in Australia reported by Palmer [7] who shows that it may take over 20 years for an off grid system to reach an ERoEI of 1:1. This is despite the fact that the panels themselves have an ERoEI of 9. Going off grid requires the system to be over-dimensioned and battery storage to be added. This has a dramatic and negative impact upon the ERoEI. Solar PV is in fact dependent upon the existence of a fossil fuel or nuclear based grid for it to work profitably. It appears that solar, like wind, is parasitic.

Preto’s number is for a broad bound analysis and based on Spain where the load factor is likely to be around 20%. Should this low number of 2.4 be correct the Scottish equivalent, adjusting for load would be 7.3/20*2.4 = 0.88. In other words, Scottish solar PV will never recover the energy used to make the panels. It is in fact worse than a waste of time. All of the CO2 emissions associated with manufacture are produced before the panel produces a single kWhr. Energy recovery may go on for 20 to 30 years and all of the CO2 associated with that promise of future energy is produced decades in advance. No wonder little progress is being made in reducing CO2 emissions.


  • It seems likely that solar photovoltaics deployed in Scotland will never repay the energy used to manufacture the panels. They will therefore produce more CO2 than if solar was not deployed at all and the emissions are emitted decades in advance of the solar electricity being produced.
  • The Dutch have effectively zero hydro power because the country is flat. Scotland should have zero solar power because the Sun rarely shines. Glasgow City Council [1] would do well to reconsider their Green Dream which is in fact Green Folly.
  • Scotland was once a proud engineering nation, a global leader in nuclear engineering, that is now being led by Green evangelists. James Watt and Lord Kelvin will be turning in their graves.

[1] The Telegraph: Revealed: how Glasgow will be Scotland’s first Solar City
[2] Euan Mearns: The Efficiency of Solar Photovoltaics
[3] BP: Statistical Review of World Energy 2013
[4] MetOffice: Historic station data
[5] Euan Mearns: The link between sunshine and temperature based on UK climate records since 1933
[6] Professor Charles Hall: The Energy Return of (Industrial) Solar – Passive Solar, PV, Wind and Hydro
[7] Graham Palmer: Energy in Australia: Peak Oil, Solar Power, and Asia’s Economic Growth

It may be hard to believe, but sunshine is going to transform Glasgow into one of the world’s greenest cities [1]

Yes, it is hard to believe since it is untrue.

This entry was posted in Bonkers, Climate change, Energy, Political commentary and tagged , , , , , , , , , , . Bookmark the permalink.

36 Responses to Solar Scotland

  1. Joe Public says:

    “To the left of the picture is a large roof section facing due south. It is inexplicable that any properly educated engineer could construct a system in this way.”

    Three points Euan:

    1. The buildings may be owned / operated by different companies / departments?

    2. There appears to be a large tree shadow to the east of the south facing roof. Maybe the tree has a preservation order and will cast shadow on the roof at the time of optimum solar intensity?

    3. Planning permission approval may have required a certain number of ‘renewables’ credits? Solar panels qualify. NB Generating credits is NOT the same as generating renewables!

    • Euan Mearns says:

      Joe, as far as I know the whole building is owned by a single medical charity. The tree shadow is a good spot. Here is another pic. The S facing roof is bathed in sunshine at the time of photograph while the E facing roof is likely edge on to the Sun and half in shadow. The tree may shade the roof for an hour or two. But I suspect your point 3 is the winner.

  2. Glen Mcmillian says:

    I agree that pv in Scotland is a bum deal- a very very poor way to spend scarce money and a waste of useful resources.

    BUT – in terms of computing the energy return on energy invested is it really useful or fair to consider the food eaten by the people in the industry in such an analysis?

    I think it matters in the very long term because perhaps it proves that pv can never be a self supporting industry capturing enough energy to bootstrap itself without the help of fossil fuels.

    But in the short to medium run- all those people involved in the industry are eating and working any way at something.Whatever that something is it may well use up more materials and energy per person involved that pv manufacture and installation.

    Materials and energy are being consumed at prodigious rates manufacturing stuff that is essentially worthless. Nobody seems to think the energy and materials used in flying to the beach in the summer and the ski resorts in the winter is wasted.

    Of course any well informed rational person living in Scotland would put the money and sweat in energy efficiency and conservation rather than pv.

    If we are to make it thru a transition away from fossil fuels to renewables then conservation and efficiency are going to be at least as important as the renewable technologies themselves.

    I strongly suspect that money spent on conservation and efficiency will pay a far better return than money spent on wind or solar or any other renewable for the easily foreseeable future in nearly every case.The energy return on energy invested will also be superior if spent on conservation rather than renewables on a small scale.

    I am personally switching to led lights as my CFL’s burn out.If I ever own a new car it will be either a plug in hybrid or pure electric.

    The solar resource is fairly good here in Virginia where I live but it will be a long long time before juice from the grid is more expensive than generating my own.

    When I finally get some pv it will be because it can be justified more as an insurance policy than for any other reason.

    I do think there is an excellent possibility pv can be deployed on a massive scale in very sunny areas and the juice moved via HVDC transmission lines to far off cities economically in the not too distant future.It seems the cost of such transmission infrastructure will be coming down as the technology is more widely adopted.

    • Euan Mearns says:

      Mac, I agree entirely that energy efficiency and conservation are and will be King. In fact there is probably a lot of good work done by government in this area in the UK. This needs to continue. But the fantasy of Desertec, that seems to have captured the imagination of the political elite this side of the pond, needs to be buried for the time being. Europe needs lots of new nuclear power, not 2000 mile long HVDC cables leading to intermittent solar energy hosted by Islamic states.

      • Glen Mcmillian says:

        I am somewhat of an old fashioned conservative in the sense that I would never enter into any sort of large scale long term relationship with anybody for any reason without giving plenty of sober thought to what might go wrong.

        And having spent many a long evening reading history while almost everybody else watched sports or other drivel on the idiot box I think you are dead on about putting such a huge investment as DESERTEC in a country dominated by actual and potential enemies.

        But this does not mean that such installations cannot be placed elsewhere on territory in the hands of people with similar cultures and values and close ties to the people on the receiving end of the long distance lines.

        A two thousand mile line would be long enough to get Arizona sun to a lot of people on the East coast of my country at the right time to help with the afternoon peak.

        It should be a rule that in discussing such issues as these that the person doing the talking mention the time frame he has in mind.

        Fukushima hit me in the gut like a bullet but in the end I have to agree that the short to medium term welfare and security of Western Europe can best be provided for with a new fleet of nukes.I think we should build new nukes here in the states as well.STARTING YESTERDAY.

        But I really doubt it will be possible to get such a fleet of new nukes permitted and financed any where in the Western world in a timely manner and so I conclude that we need to keep the pedal to the metal on renewables as well.

        In the end the real value of renewables in economic terms is going to depend on their allowing us to use less of all the fossil fuels over the next few decades and thereby slow down the inevitable increases in the costs of such fuels.

        As expensive as renewable power may be it can and will reduce the consumption of coal and natural gas to a truly substantial extent going forward and it will also reduce the consumption of oil to some extent.

        We have existing wind farms in the states that supplied over four percent of our total electricity last year. They will supply approximately that same number of gigawatt hours for the next couple of decades at least at very low costs except interest and subsidies since they are now built and therefore ” sunk” money.

        Four percent of the money spent on coal and natural gas by the electrical generating industry annually is a hell of a lot of money. Multiply that by twenty years ….

        And then add in how ever much more you think is justified as a ” bonus” for holding down the prices of coal and gas to some extent…….

        We will just have to bear the cost of maintaing a robust enough fossil fuel infrastructure to compensate for intermittency as well as bearing the cost of the wind and solar farms.

        This will without question require a new set of rules determining who pays for what in the electricity industry.The old business model cannot stand forever with more and more renewables being incorporated into the mix.

        Now here is an interesting thought about seasonal intermittency.If we take the example of Germany the solar output is very poor in winter but tolerable in summer.However much juice the Germans get in summer from their solar installations is that much they don’t have produce by burning imported gas and coal on an annual basis.

        A poor working man may budget from week to week as his paychecks arrive but a country such as Germany budgets on an annual basis.

        A Euro saved in July on coal and gas is worth as much as a Euro saved in January unless the price of coal and gas is up in the winter compared to the summer.I am not sure how much gas can be stored in the off season but coal can be stored almost indefinitely at no more expense than the interest on the purchase money so coal prices should be fairly steady for German utilities.

        I know gas prices vary considerably over the course of a year it the market place but I assume most utilities buy on contract for the most part and thus don’t pay that much more in the winter than they do in the summer.

  3. Dag Johansen says:

    Solar PV is definitely better in some places than others. But it is useful just about everywhere to some degree. In California where I live, it is great. But other places have other advantages. The Icelanders have their geothermal. And the Scots have great wind. But even in those places, it is nice to have solar PV to help balance/buffer those other sources.

    • Euan Mearns says:

      Dag, you are talking from your heart, not your head. I of course agree that sunny localities are better and may even make sense – Roger and Dave have solar panels. If Preto’s numbers are correct and you add solar tracking then you end up with ERoEI of 3.5 in low latitude sites. That means you can use one unit to make a new panel over the 30 year life (if it lasts that long) and have 2.5 units to power you, your hospital, schools, agriculture, transport etc – good luck with that. In Scotland we don’t get to use any of the power, it all needs to be saved to make 0.9 new panels to replace the old ones when they ware out – or get covered in moss which is what happens in this part of the world.

    • Roger Andrews says:

      Might add that adding PV to wind doesn’t help “balance” wind. In fact it makes the problem of matching generation to demand far worse. Geothermal, which commonly runs at load factors of around 90%, doesn’t need balancing.

  4. Roger Andrews says:


    You may complain about your cold and cloudy climate, but did you know that at latitude 57N you get 17% more incident solar radiation on June 21 than I do at latitude 20N? On December 21, however, it’s a no-contest:

    The bottom line is that you can in fact generate quite a lot of solar power in Scotland, but only in the summer when you don’t need it.

    On the question of which way to point solar panels, after watching my output fluctuate on a partly cloudy day I’ve concluded that “solar power” is a misnomer. The output is actually a function of how light the sky is. There is of course more light when the sun shines, but I still get about 50% output when the sun goes behind a cloud and the shadows disappear. This being the case it may indeed not make all that much difference which way the panels are pointing in a cloudy place like Scotland as long as it’s towards the sky. Flat may in fact be a perfectly acceptable inclination, although I’m not sufficiently in love with spherical trigonometry to attempt any quantitative calculations. Maybe Dave Rutledge can add something here.

    You may also find this Glasgow University study interesting. It goes into the question of matching solar and wind to future demand using an approach similar to the one I’ve been using and reaches the following conclusions:

    “Across the whole year, there are large periods of electricity deficit during winter and large periods of electricity surplus during summer. On individual days, fluctuations can be induced by factors such as whether the Sun is above the horizon and whether or not it is windy.”

    “The challenge for a future grid is to somehow find a way of using the surplus to cover the deficits…”

    • Euan Mearns says:

      Roger, I am in information and bio fuel overload. Can you please explain what this means.

      On individual days, fluctuations can be induced by factors such as whether the Sun is above the horizon and whether or not it is windy.

      These whizz kids at the UoGlasgow have got me here. Is this a breakthrough in understanding or what?

      • Roger Andrews says:

        Euan: What the UoG whizz kids are saying is that you don’t get any solar power when the sun is below the horizon and that the amount of wind power you get depends on how strongly the wind is blowing. These facts were previously unknown to most politicians, greens and renewable energy salesmen, so you can if you like classify it as a breakthrough in understanding. 🙂

    • Glen Mcmillian says:

      If Scotland CAN use the full output of Scottish solar installations in the summer then the country DOES need that output.

      If the country is buying imported coal and gas to generate electricity then the saving are as worth as much in dollars or Euros in July as they are in January allowing for seasonal price fluctuations in coal and gas markets and whatever the costs of load balancing may be.I have not so far been able to find out how much it actually costs in terms of extra gas consumption to handle the load balancing chore.

      There are engineering and physics problems- and there are plain old money problems.Money is about as big or a bigger problem in the short to medium term as the engineering and physics problems.

      The real problem in the short to medium term is not engineering or energy returned on energy invested. Our society wastes energy in prodigious amounts every day.Energy is not in short supply in the near to medium term. Energy and money wasted on pv in Scotland is no more wasted if spent on pv than if it is spent on heating a poorly insulated building.

      Money is in short supply NOW.

      In the long term energy truly will be in short supply and then engineering and physics will trump money.

  5. burnsider says:

    As I live even further north than gloomy Aberdeen (Caithness), I couldn’t agree more with your post.

    It is a complete misnomer to call solar PV (or wind) power ‘renewable’ unless, using the electric power alone from the first solar panel (or wind turbine), another one can be built. This exercise would have to include mining the iron ore, bauxite, etc, shipping these to the UK, processing the ores into metals, etc, making the solar panels/turbines, shipping them to their final destination and installing them. Such an exercise done entirely with ‘renewable’ electric power and no input of fossil fuels is pretty much impossible, so ‘renewables’ do not exist and likely never will.

    My neighbour installed solar panels (on the ground, ideally aligned due south at about 23 degrees above the horizon) just in time to benefit from the higher feed-in tariff before it was reduced a couple of years ago. I was subsequently invited to take a look inside his shed at the inverter assembly one bright and sunny day and the powermeter was showing 3.3kW, which is quite commendable. While I was watching, a small cloud passed over the sun and although the ambient light didn’t look much dimmer to my eye, the power dropped immediately to 600W. I haven’t seen the power meter on a nice gloomy day in December…

    Finally, worth remembering that solar power worldwide starts off with a load factor of 50%, due to the inconvenient onset of night on a regular basis (and just when you want to turn on the lights)

  6. Alister Hamilton says:

    Hi Euan,

    You might like to play around with this

    Best wishes,


    • Euan Mearns says:

      Alister, that’s very cool. The annual average for Aberdeen is 10% efficient for a S facing panel, 7.6% for E facing and 4.6% for N facing. All done at 42˚ slope. Curiously, choosing the optimum slope for E and N facing it chose 0˚, i.e. the system laid the panels flat. And so I am now left wondering if the measured efficiency for the UK less than 9% is down to poorly orientated panels. Faro in the S of Portugal comes out at 18%.

      • Roger Andrews says:


        You’re “left wondering if the measured efficiency for the UK less than 9% is down to poorly orientated panels.”

        The impact of panel orientation is one of the things I’m trying to get a line on here, and slowly I make progress. In the case of your question the answer is no, the low UK load factor isn’t a result of poorly oriented panels.

        Here’s a graph of quarterly UK solar PV load factors between 4Q2011 and 4Q2013. The average over the period was 8.79% – very close to your number. The data, however, take into account only installations of over 10KW, and almost all of the total capacity above 10KW is made up of installations larger than 500KW. I think we can assume that installations this large will have the panels pointed in the right direction, and if so then a ~9% load factor is about the best you can expect to get in UK.

        But why is this only half the solar PV load factor in Portugal and Spain? Latitude will have some impact, but the main reason is that the UK is over twice as cloudy:

        Finally, bearing in mind that UK solar farms are heavily concentrated in the sunny south of England, what load factor might you expect to get in Aberdeen? My guess would be about 6% if you get all the panels pointing in the right direction (whatever that is) about 5% if you don’t, and roughly twice that if you could figure out how to make the clouds go away 😉

    • Euan Mearns says:

      Choosing the 2 axis tracking option I get 13.1% efficiency for Aberdeen, that is 31% more efficient than fixed.
      Shetland is 8.2% for fixed S facing and Bournemouth is 12%. All the numbers hang together pretty well with my post.

      • Alister Hamilton says:

        Whitfield Solar use a solar tracking and concentrating system

        based on high efficiency PV cells (>40%). Perhaps they are Spectrolab cells?

        I have a feeling that Rodger Bentley (U of Reading) has had something to do with Whitfield Solar (although I could be wrong).

        Perhaps you could ask for a guest post?

        I’d be very interested!

        Best wishes,


      • Roger Andrews says:

        Eaun: You’re going to have to send me instructions on how to use that calculator because all I can get out of it is zeros. 🙁

        However, I suspect it doesn’t allow for cloudiness, and if so your results could be misleading.

        Today it’s just like Scotland here – ten-tenths cover – and my panels are pointing away from the sun (which is low in the eastern sky) straight at a thick black cloud. But they’re still delivering 100W. How much would I be getting if there were no clouds? Maybe 500W. And yesterday afternoon when it was partly cloudy I was getting about 50% of maximum (~800 of ~1600W) when the sun went behind a cloud. And even as I write the thick black cloud has been replaced by a less thick gray cloud and output is up to 500W.

        Based on these results I’m going to assume that I get about 2.5 times as much solar power when the sun is shining as I do when it isn’t.

        Now the median cloudiness in Aberdeen is ~84%. It doesn’t vary much from month to month and I suspect not much from day to night either:

        With 84% cloudiness and applying my empirical 2.5:1 sunny/cloudy generation ratio, I find that roughly two-thirds of the solar power generated in Aberdeen will be coming from cloud-diffused rather than direct sunlight. And the main controlling factor with cloud-diffused light, of course, is how thick the clouds are, not where the sun is. The brightest clouds may be – often are – where the sun isn’t.

        Assuming this simplistic analysis is anywhere near correct, the implications are:

        * It will be very difficult if not impossible to determine an optimum panel azimuth and inclination that takes both solar radiation and the impact of variable cloud cover into account.

        * However, you can’t go too far wrong by pointing your panels straight upwards.

        * Cloud cover in Aberdeen cuts clear-sky recoverable solar energy by a factor of about two.

        Is there any chance you could get some actual production numbers from a local solar company? Or are they kept a closely-guarded secret to avoid alarming the populace? 😉

        (The black cloud is back, Output now down to 200W.)

        • Euan Mearns says:

          Roger, you should check out the radiation data base link. They have regional irradiance models based on data. Its pretty easy to use. I left all options on default. Checked the Also Optimize Azimuth button. Pressed calculate with web page output. Before doing that you need to select your location by entering it in box above map or simply clicking on map. This only does Europe and Africa. Selecting a spot on the coast of Mauritania which may be analogous to your W coast location in Mexico I get 3.76 kWh per day average for 1 year from a 1Kw panel = 3.76/24 = 15.7% load. Have to assume that Mauritania is cloudier than Portugal.

  7. Mike Booth says:

    Thanks for the illuminating blog!
    If you haven’t already come across him, Tom Murphy has been doing a lot of ‘brass tacks’ physics on house insulation, solar performance etc based on his own experience. All well documented at
    His observations make illuminating, and sometimes amusing reading.

    BTW, the specified lifetime of a PV panel is to 80% of as-new output. Assuming the panel physically survives that long (BIG assumption) it will continue to produce at an ever reducing output which leaves us to make a choice of when the EROI integration will be wrapped up. When the 4kWp installation output falls to 3kWp, or 2kWp, or….. After all, by then it has paid for itself a long time ago.

    Re generation under diffuse illumination, PV panels perform surprisingly well with light cloud cover, and the output is thereby enhanced at times when they would not be in direct illumination from the sun. Similarly with reflections from broken cloud. So not all is lost on a cloudy day.


  8. Nigel Wakefield says:

    Just to add a little here to the understanding of solar PV.

    A panel’s rated capacity is based on output with irradiance at 1000 watts/square metre, and with panel surface temperature at 25 degrees C.

    1000 w/sqm is bright sunshine, so under those conditions the panel surface, in most areas, will heat up rather quickly. PV panels have a temperature coefficient, which measures the decrease or increase in output at 1000 w/sqm for very degree C above or below 25C for the panel surface. Temperature coefficients vary per panel and manufacturer, in a broad range from 0.30 (market leading) to 0.50 (poor performance, but usually cheaper panels).

    A reasonable industry average is 0.42, which means that for every degree difference in panel surface temperature from 25C, the output will decrease or increase by 0.42% (at 1000 w/sqm irradiance).

    Wind can be quite an important factor in solar output, since a breeze moving across the panel surface will help to cool it down thereby increasing output. This probably explains why output in Portugal (with prevailing westerlies off the Atlantic ocean) is better than Mauritania (where much of the wind is easterlies from the Sahara). Panel surface temperatures (under the same irradiance) are likely to be higher in Mauritania than Portugal, thereby reducing output. I’d guess it’s windier more of the time in Portugal than Mauritania.

    From observations of my own PV array, I typically get the best instantaneous output on cold clear days from March to April. A cold, breezy sunny day in March, with temperatures in low single digits celsius and the sun only available 12 hours and still relatively low in the shy (basically equinox) will yield nearly as much output as a sunny, calm day in June at the solstice, despite higher irradiance and longer days. This is because the panels may be as much as 35C hotter in June than in March, leading to a 10% drop in output efficiency (I have Panasonic HIT panels which have quite a low [ie good] temperature coefficent).

    This is all a bit “layman knowledge” but I hope it adds a little to the thread….

    • Euan Mearns says:

      Nigel, thanks very much for a most informative post. I have been a bit loose in my usage of the term efficiency. There are two main factors at work 1) the physical efficiency of the panel and 2) load factors controlled by the site. I’m betting my estimate aggregate efficiency / load of 7% for Scotland is not too wide of the mark and that means we would have to install 14 kW of panels to produce on average 1 kWh per hour over a year. That would cost €28,000 in up front costs to keep a 1kW fire burning 24/7 for 30 years. I have to pay €1000 now for energy I may or may not use in 30 years time.

      • Nigel Wakefield says:


        You’re probably about right in your estimation of average load factor for PV in Scotland. Down here in Kent (SE England for our overseas readers) I get a load factor of about 9.3%, though I’d have to say my panels are not optimally positioned due to the nature of my roof. 50% are SE facing, 50% are SW which means my “load duration” curve is longer than 100% due south would be, but I do suffer from some shading issues early and late in the day.

        Your estimate of 28 kW requirement for an average output of 1 kWh/hour is really not very useful. There’d be days in winter when you got nothing at all (indeed I’ve had zero output days in May and October previously) and you’d generate far more than you could usefully use in summer.

        Without storage, PV is simply a useful way to reduce reliance on grid electricity. In common with many other owners of PV installations, I have modified my consumption patterns to maximise my use of PV generated power. I tend to save tasks like vacuuming, washing clothes, etc for times when the sun is shining. Spare output goes to an Immersun device which automatically heats the water in my cylinder via the immersion element, etc.

        My total usage of grid generated electricity is less than 1,350 kWh/year, half of which is on off-peak Economy 7 rates for electric heating of water and running appliances with timers, etc mostly in winter when my PV output is low.

        I estimate that I could go completely off-grid with 50 kWh (perhaps as low as 30 kWh) of storage, though this remains a completely uneconomic proposition for the time being. I’m interested to see developments in storage technology from Elon Musk, who hopefully can leverage his ownership of Tesla Motors with his shareholding in SolarCity to bring battery efficiencies up and costs down – time will tell.

        • Roger Andrews says:


          My 2.25 KW solar system here in sun-drenched Mexico is connected to the grid so that power flows out when the system generates more power than I’m using and in when I’m using more power than the system is generating. At the end of the month I pay for total power in-out, and for the last few months my bills have been zero because the system has been generating more electricity than I consume. (I get credited for the surplus when consumption begins to exceed output again, as it will later this year).

          With this arrangement, however, I have no incentive to modify my consumption patterns to match my solar output, as you do, which prompts me to ask you what your incentive is (I assume your system is interconnected with the grid). Is it pricing, or maximizing energy efficiency (I have no problem with that) or some combination of factors?

          Your 9.3% load factor sounds about right. It’s slightly higher than the 8.8% UK average, maybe because you get a little more sun in Kent?. But it’s still less than half the average solar PV load factor in Spain, and as I argue elsewhere on this thread only a small fraction of this difference can be attributed to latitude and misaligned panels. Most of it occurs simply because the UK is twice as cloudy as Spain, and clouds and solar power don’t mix too well.

        • Roger Andrews says:

          Nigel. Re your initial comment, I suspect that temperature may be the reason my solar panels operate at only 20% load factor. At latitude 20N and 1500m elevation they should do better than that, right? But we don’t get much wind here and the panels do get very hot. I could of course cool them down by directing a fan on them but the EROI would probably be a little on the low side 😉

        • Euan Mearns says:

          Nigel, its a fascinating discussion, wish I had time to be more involved but I’m plotting charts – Nigeria! The arguments against owning PV in the UK are as follows: 1) like wind it is parasitic, it is undermining the viability of the grid upon which it depends to be viable, if everyone did it the grid and the economy would collapse 2) it is enabled by feed in tariffs that everyone who doesn’t have PV have to pay for, normally poor folks and 3) if the EROI numbers quoted by Pedro and Graham are correct then the reason the UK government have created the market mechanisms in favour of solar are totally undermined.

          The low EROI numbers effectively take into account the energy cost of intermittency. You say yourself that you cannot afford to pay for batteries which would be placing that cost on you. But the way the system is rigged, you are allowed to have that cost you cannot afford to be paid for by everyone else.

          No sleight intended, but these renewables schemes are dreamt up by City boys who are accustomed to bleeding the public to line their own pockets.

        • Without storage, PV is simply a useful way to reduce reliance on grid electricity

          I think this highlights the problem of grid-connected PV and how we pay for the reliability service that the grid is providing.

          When we buy electricity, we are really mostly paying for reliability but we are charged according to energy. In the old days before PV and few people had air conditioning, the costs of providing the reliability to each consumer and their energy use were aligned closely enough that the mechanical accumulation meters sufficed. Electricity was deemed an essential service and necessary for development and there was an informal social contract that those that used the most energy paid the most. We essentially had socialized tariff pricing.

          But grid-connect PV, and the growing use of A/C and thrown a bolt in the sprocket. PV reduces the householder’s net energy but the cost of providing the reliability hasn’t changed. So effectively homes with PV (and A/C) are being cross-subsidised by everyone else, or essentially “gaming” the tariff structures. This is a big issue in Australia, and actually, A/C is the bigger problem, but is raises the issue of how we should be allocating costs. Demand-based tariffs, time-of-use tariffs, clever demand management, these are difficult problems without a simple solution.

  9. Ian Chisholm says:

    Interesting info from Nigel on the output of PVs relative to ambient temperatures and seem to indicate in cold climes such as Scotland there would be a gain. In the preceding posts I did not glean any info on the duty cycle of output over a year and comparing high to low latitudes. I do remember an old friend of mine, now sadly passed away, Dr Kerr MacGregor of Napier University Edinburgh, telling me Scotland was a good location for PV installations because of our long summer nights and that the diffused light we had was not a huge disadvantage. I hope I haven’t taken Kerrs’ words out of context but this was at a time when PVs were in their infancy. As to HVDC grids…..these are essential and should be a priority construction across the EU and probably the USA. Whatever form of generation is used the grid will efficiently transfer load/demand. Euan seems to make a case for Nuclear which neglects the social issues making it an unlikely contributor in the longer term. The new arrangements for the Hinkley Point plants include decommissioning costs but not long term storage of waste….the public will soon have to pick up the bill for the decommissioning of Sellafied then all of the aging Nuclear power stations. The costs will run into many billions of pounds….a political hot potato mitigating against new installations. It is against that background that the use of PVs should be weighed. Whatever upfront costs of renewable are the fact is the fuel is free and near infinite. I think the whole basis of costings shown for the manufacturing PVs is very limited and fails to take into account the employment benefits of manufacture. The siting of huge arrays on unproductive land also makes a case for PVs. and HVDC grids. The point raised by someone about the risks of reliance on unstable Muslim States is not valid….who can say which States will be stable in a decade….look at Gas supplies fro Russia….regarded as stable six months ago.

    • I think the whole basis of costings shown for the manufacturing PVs is very limited and fails to take into account the employment benefits of manufacture.

      The idea of employment sounds good in theory, but if the product is costing more, requires more labour, and does not add to the net energy available, is it a net benefit or a net burden? When a farmer buys a motorised harvester to replace labour, productivity improves, costs lower, and society can complexify and spend less on food. Agriculture in Cambodia is still about 70% of the workforce but only a few percent in developed countries. Cambodians don’t have the opportunity to work in one of the thousands of occupations that people can in the UK, where food is only a minor part of expenditure. If we deliberately reverse the productivity growth of our energy systems, what will be the impact? Lowering economy-wide productivity, net-employment contraction, declining living standards etc.

  10. Roger Andrews says:

    Barack Obama is installing solar panels on the White House roof.

    The panels are being installed flat, or as close to flat as makes no difference 🙂

    No tracking capability either.

  11. Roger Andrews says:


    I’ve done some approximate economic analyses of domestic roof panel solar installations in England, Aberdeen and Mexico (my system) and thought you might find the results interesting. They are of course subject to uncertainty but should be broadly indicative:

    * Rooftop solar panels in England give an IRR of 4.5% and payback in 17 years with subsidy payments and an IRR of minus 13.4% (i.e. no payback ever) without subsidy payments.

    * Rooftop solar panels in Aberdeen give an IRR of 0.9% and payback in 23 years with subsidy payments and an IRR of minus 15.6% without subsidy payments.

    * My system gives an IRR of 22% and payback in ~5 years without subsidy payments.

    Assumptions supplied on request.

    • Euan Mearns says:

      Roger, sounds good and makes sense. Do you fancy doing a short joint post on this including the numbers – installations costs and FITs etc. Could be good to have Dave R’s and Nigel’s real systems included. I have a refined way of getting the load factors that corrects for current year production never being a full year for the year end capacity – applied this to wind and it seems to work a treat. So if I did that for solar, we could perhaps stick that at the beginning of the post.

      The conclusion that UK electricity consumers are subsidising loss making ventures and enabling the practitioners to make money is rather important.

  12. I think the whole basis of costings shown for the manufacturing PVs is very limited and fails to take into account the employment benefits of manufacture.

    The idea of employment sounds good in theory what value or service is PV providing society and what does it cost? When a farmer buys a motorised harvester to replace labour, his productivity rises, costs lower, and society spends less on food. Cambodia has 70% of its workforce in agriculture, the developed countries a couple of percent, leaving the rest of society to complexify. Cambodians don’t have the opportunity to work in any of the thousands of occupations available in the UK, nor all the education and healthcare. Energy systems are the same. We could engineer this backwards to artificially increase energy industry labour, but subsequently reduce national productivity and contract other industry sectors, increase energy costs, reduce tax receipts and reduce living standards. If PV increased energy productivity, it would already be happening and energy would be getting cheaper.

Comments are closed.