Exploring Solar P.V. - Fact versus Fiction

[Mitch]

Exploring the vagaries of Solar P.V
Separating fact from fiction
By Mitch Smith – April 2009.

In this article, we explore my personal journey through the world of Solar Photo-Voltaic electricity generation. The “fiction” part is mostly conjecture and un-supported information obtained via the Internet and friends – the “fact” portion is real life experience of my own.
Let’s begin with the background which led me to spend £11,500.00 on Solar P.V. for my live-aboard Narrow Boat. Soma is a 40 ft craft with a 26ft cabin area. I have lived aboard since 2001 and, in 2003/4, decided to turn her into as close to a “flat” as possible. She has her own water filtration system – from canal water to drinking water quality, and all the standard, 240v, domestic appliances, including 240v lighting. Since she is gas-free, I have electric cooking in addition to the washer/drier, dishwasher, separate fridge and freezer etc. All this requires, on average, 9.7 Kw/hrs of electricity per day. Let’s call it 10 Kw/Hrs per day for simplicity. These are genuine measurements, taken over a week, and then a month, with the same results. I was very fortunate in that my workplace was situated canal-side, and for years I lived virtually “in the parking lot” outside work. I also had a “free” 13 amp shore-power hook-up provided by my employers. Soma also has a 2000 amp/hour battery bank and 3 Kw inverter – a 13 amp plug point gives just on 3 Kw, so I could live on or off grid with no discernable difference to life-style. In May 2008 everything changed – redundancy. Not only did my income reduce to zero, but I lost the shore-power hook-up as well. I did receive an acceptable redundancy payout, but figured that if I spent it on generators, diesel/petrol etc, it would soon be gone and I would be back to square one – no electricity. I began an in depth investigation as to how to provide the required energy long-term. Although somewhat convincing, “free energy” machines didn’t quite cut it enough for me to give my cash to a stranger, wind seemed a little too erratic, but the sun rises every day without fail – sure it’s overcast and cloudy sometimes, but it definitely get’s light and dark regularly every 24 hours, so we settled on Solar P.V. as the main source. I thought, wrongly, that I was being somewhat conservative in calculating on 6 hours a day of sunlight in Summer and around half that over the Winter months. Given space, (and financial), restrictions, I could get a 1.450 Kw panel array on the roof. Again, let’s call that 1.5 Kw for simplicity. A quick calculation says I would get around 9 Kw/hrs per day in Summer and 4.5 Kw/hrs a day over Winter, requiring me to run the engine a few hours a day over Winter, but providing around 90% of my requirement in Summer. This was not looking bad at all, so 21 x 70 watt panels were duly purchased, along with a 60 amp regulator and cartons of glue and really thick cable. Sure, the panels are the semi-flexible, unbreakable, glue-on, walk-on variety, so they cost a little more than your standard glass version. Four days of gluing, drilling and wiring first week of July 2008, and we settle back to enjoy all this fuel-less energy! Umm, NOT! Within a week the batteries are flat, and I haven’t even thought of using the washing machine or dishwasher yet! The fridge and freezer die and food gets spoilt – horrors. Must be something terribly wrong, so – to imitate our leaders – I begin an “enquiry”. Herewith the result.

Fiction

Let’s first try to find out how Solar panels are measured. In the “old day’s” of the 70’s and 80’s, when I was using solar to supply remote radio repeater stations on mountain tops, a 36 watt panel actually delivered 3 amps into 12 volts under normal sunlight conditions – new technology has changed all that, it seems. We will take my “70 watt” panels as an example.
In order to “measure” a solar P.V. panel, we need a base line from which to work. The industry seems to have settled on “1000 watts of sunlight falling on one square metre” as the starting point. Therefore, if your panel is exactly 1 square metre, and you have exactly 1000 watts of sunlight falling on it, and you get 120 watts out of it, then your panel is 12% efficient. Right, so I want to manufacture solar panels, and I want to test each one for quality control, so I need a precise 1000 watt light source. The only way to do this is too create an artificial “sun” in the laboratory, which gives the required 1000 watts per square metre. It’s obviously fairly complex, and expensive, to accurately reproduce all the various wavelengths of light, ranging from Infra-Red, through visible light to Ultra-Violet and the X, Gamma and whatever else the sun produces, in accurate quantities. It’s also a complete waste of time and effort, since my solar panel can only convert a specific band of these wavelengths to electricity – it simply ignores the rest. So, we may as well only replicate the specific wavelength’s that my panel can use – much easier. While we are at it, why bother with trying to replicate the precise portion of the 1000 watts per square metre that these wavelengths make up? Oh hell, we’ll just provide 1000 watts per square metre of the right stuff – doesn’t really matter does it? Now that we have our lab “sun”, we shove our panel under it to measure it. The first thing we do is place an amp meter directly across the output – a virtual short circuit, and read the current. This will show us the absolute maximum the panel can put out – in my case 4.1 amps. Next, we connect a rheostat, or variable load resistance across the output, with our amp meter in series, and a volt meter across the output. We slowly decrease the resistance while watching the amp meter, and stop precisely when the amp meter shows us 4.1 amps. No point in decreasing the resistance any further, 4.1 amps is the maximum we can get out of the panel, and so we now read the voltage – again in my case - 17.6 volts. Now, using Ohms Law – Watts = Volts x Amps, we get 4.1 x 17.6 = 72.16 watts. Yes folks, we have a genuine 70 watt Solar Panel!!!! Out comes the advertising bumpf, brochures and all the rest – 70 watt Solar P.V. Panels!

Fact

Now let’s take our “70 watt” panel and put it on the roof of my boat, under the REAL sun. First up, we make sure that it’s a cloud free day, at 12 o’clock, midsummer – just to make sure we get the 1000 watts per square metre. Unfortunately, the real sun doesn’t give a turkey for what wavelengths of light my panel wants – it just delivers the same quantities as it has done for millennia, somewhat less than the 1000 watts of the frequencies my panel needs! So out with the off-the-shelf, mid-quality, amp meter and we short it across the output – hmm, 3.6 amps now....
Oh well, let’s look for the required 17.6 volt battery bank, or fridge, or telly, or radio, or lights – would you believe that NOTHING, absolutely nothing, can be found? Loads of 12, 24, 36 and 48 volt stuff, but no 17.6 volt stuff! So we settle on 12v, since 17.6 won’t charge 24 volts, but will bust a gut charging 12v at maximum current. So, back to good old Mr. Ohm, 3.6 amps x 12 volts = 43.2 watts. Ahem, my “70 watt” Solar Panel has just become a 40 watt solar panel!!!!! And I haven’t done a thing to it! Man, that’s about half of what I calculated on – that’s now looking like a bit of a poor show, will still have to run the engine every day – but the cost of the panels hasn’t changed!!
Now for the six hours of sunlight every day in summer.
The panels I bought are the “low-light” versions – “even in overcast or cloudy conditions, you will still get something out of them”. That particular statement is fiction – the REAL statement reads “even the slightest cloud or shaded sun will reduce the output 90%” – yes, that’s right, 90%!!! During the summer – July, August 2008 – I got an average of one sunny day a week in London NW10. On those days I averaged around 60 to 75 amps charge at midday – for the other six, cloudy, days I averaged around 6 to 7.5 amps charge rate at midday!! My carefully researched and calculated 9 Kw/ hrs a day from a 1.5Kw panel array has turned out to be just under 1 Kw/hr per day average in summer. If I had any idea of the “fact” behind the “fiction”, I would never have parted with the over eleven thousand pounds – of that you can be quite sure!!
When calculating Solar P.V. requirements , or expected outputs, the first rule is to divide any advertised ratings in half – treat a 120 watt panel as a 60 watt panel, then, if you want to use it in London, U.K., divide again by 10 – using that formula, you should get a little more than expected. In my case, in order to provide my 10 kw/hrs per day all year round, I would need about 400 “70 watt” panels at a cost of about £220,000.00.

You have been warned!

Oh, anyone want to swap a load of solar panels for a steam engine with a few alternators bolted to it? Thought not......

[snow hope]

Thanks for your experiences Mitch. That does all sound pretty disappointing. Maybe it is no surprise that it is countries like Spain and Greece who seem to be using PV in a bigger way? Don't forget that 2008 was a crap summer in England. 2009 could be a good bit better - 2 or 3 full days of sun/week, but your results are still eye-opening..... Let us know if it picks up this summer.

I had already decided that PV is not for me up here in the rain-soaked N Ireland. I had others tell me that I was wrong, but fortunately I am old enough and wise enough to have not believed them! Mind you it is still very disappointing how little electricity you are getting Mitch.

[adam2]

I am very sorry to hear of your unfortunate experience.

IMHO some unduly optimistic assumptions were made in desigining your system.
In the absence of more detailed data, I generaly assume 1 hour a day of peak sun in the winter, and 3 hours in the summer, as a fair average for UK use.
I would therefore expect a 1,500 watt array to produce about 1.5KWH a day in winter and about 4.5KWH daily in summer.
And that is for roof mounting at an optimum angle, if mounted on a boat I would expect an additional reduction to perhaps two/thirds of the above values, owing to the non optimum angle.
Result perhaps 1KWH a day in winter and say 3KWH in summer.

Battery charging brings additional losses since the voltage of the PV modules is normally in excess of the battery voltage, and this excess voltage is wasted (unless a maximum power point tracking charge controller is used)
In the absence of more detailed information I would assume a battery efficiency of about 90%, and an inverter efficiency of about 95%.

Also electricity that has been stored in a battery bank is suprisingly expensive, even if the input power is free.
Consider a typical 12 volt deep cycle battery that costs £50 and stores 0.5 KWH at 50% depth of discharge.
The battery life is somwhat variable, but 200 cycles might be a fair assumption, therefore over its life time the battery has cycled about 100KWH, at a cost of £50. Thats 50 pence a KWH, or more than grid power, and thats without allowing for the cost of making the power in the first place.
Of course under good conditions the battery might last 400 cycles which would halve the cost per unit to 25p/kwh, which is still more expensive than most grid tarrifs.

To generate 10 KWH a day in winter would require an array approaching 20 kilowatts peak, which would be a vast investment and would not fit on a boat.

[JohnB]

Surely trying to live a conventional land based lifestyle with cookers, washing machines and other standard domestic appliances is a big part of the problem. With a big PV array and a big wind turbine in an optimal location, it would be quite difficult and expensive, but on a boat with the PV mounted flat of the roof, and no wind turbine, I would have thought it would be impossible.

I've seen how much power I've missed out on, having my one 60w PV module mounted flat on my van roof. I survive on low powered 12 volt equipment, and I have the advantage of the battery getting a good charge when I'm travelling.

[IanG]

Soory to read your tale mitch. Although a few more paragraphs might make it a bit easier to read.

From memory, you used a flexable panel, which you stuck down on the roof of your boat?

If so, I suspect you paid a preimum in terms of price per watt and I would imagine that its performance wouldn't be optimal.

The following it data from my 16 * Sharp NU 180 array. 2.880 kWp system. Its performing very well and if August hadn't been so bad, I've have already reached the predicted annual output.

[RenewableCandy]

I think the point about 2008 being a crap summer was spot on: the sun hours for Aug 2008 were about 20% down on the UK average. Also a lot of figures assume you've put our panel at the optimum angle (South-facing, 40deg of elevation) whereas yours are horizontal, which would knock about 20% off your output.

You might have to kick the drier into touch, mind.

[adam2]

As others point out, powering large conventional energy hungry domestic appliances from a PV array is unlikely to be viable.
It is not actually imposible, just unreasonably expensive.

Here is how I would size a PV system to supply 10KWH a day in winter.

After allowing for battery and inverter losses, 10 KWH a day output, will require the production of about 11KWH.

Presuming one hour of peak sun a day in winter, this suggests an array of about 11KW, if optimally installed, or about 16/17KW if installed on a boat.

For such a large system, I would consider a battery voltage of 120 volts, which will slightly reduce the losses in the inverter.
It might also be posible to work some loads at 120 volts DC.

Such a large system would justify the use of a maximium power point tracking (MPPT) charge controller which would make efficient use of the PV output.

Such a system is entirely possible but is most unlikely to make economic sense.

In many cases, a much smaller system for lights, electronic appliances and high efficiency refrigeration would be a better choice.
Heating and cooking would be better served by wood, or if not available, coal, oil, or LPG.
A large wind turbine would be worth considering on land, but I dont believe that a big enough turbine could be mounted on a boat.
If the system is sized to power lighting, tv, PC etc in winter, then in summer it should produce a surplus which would allow limited use of kettles, microwave ovens, slow cookers etc.

[IanG]

If you look at the numbers in my table you'll see that the lows in november / december are near zero.

A bad week of heavy rain and you'll get next to nothing no matter how big the array is.

Frankly, you need an energy mix. I'm sure you'll get a near inverse of wind. Bit of PV, bit of wind and a small generator for use when the renewables isn't enough.

Sizing a PV array just for winter doesn't make sense to me.

[hardworkinghippy]

Sizing a PV array just for winter doesn't make sense to me.

I agree that a system needs to be balanced with another source of energy. Even here in SW France we've days in the winter when it's only the 2 little Rutland 910s which are charging the batteries. They're used a lot on boats.

I built our little system up by trial and error and knew what to expect each time I added on a bit more wind or another few solar panels. That's probably the reason I'm not disappointed.

It's sad to hear that you are though Mitch - especially after spending so much money.

[sjaglin]

Mitch,

Really sorry about your misfortune but I will echo what others have mention further up, I think the preparation, including some simulations, is important if you buy a system and compulsory if you build your own.

I started but testing a 12 volts system (13 watts), then I increased to 55 watts. The simulations ( PVSYST ) were correct but the battery side of things was not great for me.

I therefore opted for grid connected system, installed by a company that I wanted to be local. I got 4 quotes of 4 different local PV installer and started studying their quotes carefully. Then I made my choice, the whole thing took only 2 months but was well worth it.

I now have a 2.1 kWh system which works like PVSYST predicted (error 3-4%) and the investment seems to be even worth it (also getting all my money back was not my prior concern) now that the feed-in-tariff seems to be coming in the uk in April 2010!

Stef

[adam2]

If one wishes to reduce ones carbon footfrint, or to reduce expenditure on electricity, then a grid tied system is the way to go.
The PV modules are allways worked under optimum conditions and there are no battery losses, and of course no battery to purchase and maintain.

However in general grid tied systems do not give any protection against power failures.
They are therefore not the best choice if you fear TEOTWAWKI.

A small battery based system is IMHO an excellent idea for essiential lighting and other low power uses.
Such as system should be considered as a disaster prep and not primarily as means of reducing ones electricity bill.

[Janco2]

A very interesting thread, especially the informative data.

We have had a grid connected 6kW wind turbine for 2 years and although this generates twice as much electricity as we use, we still find we need to import half of what we use due to the intermittent nature of wind.

Consequently we have been trying to add a 2kW array of solar PV to our system in order to reduce our import from the grid.

After originally being given a verbal go ahead by Western Power, this has now been retracted
'due to length of ug service cable from local transformer'
'source impedence is circa 0.48ohms thus 6+2kW of export would create 32A x 0.48ohms = circa 16V rise and probably trip G83OV protection'
'To reduce source impedence would need existing ug service cable to be overlaid with a larger cable at a cost of £6,600+VAT' to be paid by us!

A data logger (recording voltmeter) has now been installed on our premises to record what is happening now.

We are not in a position to pay out this extra cost.
Has anyone any ideas as to how to proceed. Could it be worthwhile having a bank of batteries with less solar PV? If so how can this be used in a power failure? Would we need a separate system?
Any other suggestions or thoughts would be appreciated.

[DominicJ]

Electricity is confusing (ohms and such).

It depends how intermittent your wind supply is.
If you generate double your yearly consumption in December, but no other power, a battery isnt going to work, if its 10 minutes on 10 minutes off, a battery would do fine.

I'm quite sure you can (but dont know how) set up different circuits to run only under certain conditions, so your clothes dryer, immersion heater ect would switch on and off with the wind, in which case batteries and solar PV might work.

I'm sure someone with better knowledge than I will be along shortly.

[Ted]

Well, Western Power's calculations are correct, as they have to design for the worst possible (from their point of view) scenario even if the real likelihood of you exporting 32A is maybe less than 1%.

But I presume they are currently quite happy with your export of potentially upto 25A from the turbine.

There are two alternative solutions to keep the new system export at 25A max.

Either the PV system could be setup to only export in the situation where the wind turbine was not producing more than 4kW with diversion to a dump load, but this would obviously have to be done in a manner that WP were happy with, or to keep the PV on a battery/inverter system to run specific circuits, rather than grid-tied, which would probably be the cheaper option.

[Bandidoz]

So we settle on 12v, since 17.6 won’t charge 24 volts, but will bust a gut charging 12v at maximum current. So, back to good old Mr. Ohm, 3.6 amps x 12 volts = 43.2 watts. Ahem, my “70 watt” Solar Panel has just become a 40 watt solar panel!!!!! And I haven’t done a thing to it!

It's not the fault of the PV panel if you're going to use it inefficiently (use a switchmode converter!!! - "maximum power point tracking charge controller" as Adam2 put it). It's like operating a radio on bad SWR and complaining about its consequential poor performance.