21 November, 2015

Time To Get Solar Again

It has been quite some time since the last posts on the grand Solar Installation, July 2nd and July 23rd, to be exact!  Lots of reasons for the delay, such as trying to keep our beach from washing away (improving daily after massive efforts for 8 weeks), good computers gone bad (resolved, for the time being), lousy internet connection (OK at the moment), travel for Dennis (routine medical in the US), travel for Wilma (family holidays in the US), and riding herd on the construction crew (ongoing).  But right this minute, while the plates are spinning nicely on the poles, we will get Dennis's 3rd installment of the solar installation posted. 

Welcome back my friends to the solar show that never ends. We’re so glad you could attend. Come inside; come inside.  So far we have 1) ascertained how to orient the roof for mounting solar panels, 2) estimated how much electricity we use and need to replace each day, and 3) how large a battery bank to set up.

In this installment, we decide on which solar panel to use and how many.  In the final installment, we will get the Pro’s in to install the solar power system!

Which Solar Panel?  There are a few ways to approach this.  You could let a sales person steer you in the direction they want or you could perform your own due diligence.  I decided on the latter, despite not knowing everything I needed to know.  I looked over a variety of vendors and decided I did not want any cheap junk solar panels that have flooded the market the past several years.  A 20 year warranty on a solar panel does not mean a whole lot if the vendor goes out of business or if the panel fails and your system will be underpowered for months while trying to get the warranty claim settled while living off-grid in a “developing” country.  Here is a website for comparison of solar panels, but it is not all inclusive.  I narrowed down the list of numerous vendors based on quality and reliability.  I also wanted high wattage panels.  Here is a more thorough listing of what characteristics and parameters that should be considered for purchasing panels, found after the fact, but useful nonetheless.  Wished I had found this when I was looking for the panels.  

The vendors I considered were Kyocera, who has been in the solar business for years; Canadian Solar, which has an excellent reputation; Helios, an American manufacturer of high power high quality panels; and Solar World, an international company with US manufacturing producing excellent quality panels, but a bit pricey.  Helios had the highest wattage (300W) and best quality panels, but they went out of business due to alleged unfair marketing practices.  This occurred just before I was going to order the panels!  However, Real Goods, a “Green” vendor in California, had a sale on Solar World 275 watt panels.  Bingo!  I ordered a pallet-load of 30 panels at a greatly reduced price.  With what we ended up doing here, we would have 9 panels atop the Caretaker Cabana, 15 panels on the Main Cabana, and 6 panels for other use including as spares in case of damage to the panels by a falling coconut or tropical almond fruit (while I may be a wise-ass at times, I am not being facetious here), or loss from high winds. 
Potential solar panel destroying missiles and canon balls ranging from lemon-sized tropical almond fruits to rugby ball-sized coconuts.
We take our lessons where we can, especially if we can avoid learning things the hard way.  Our neighbors, the Harrises, lost half of their high quality BP solar panels due to 70 mph straight-line wind one night. They found their missing panels in the morning, in the trees behind their house. They could not immediately replace the lost panels with BP panels since these are not vended locally, and so had to use lower quality panels available in Belize. The restored system had lower efficiency for capturing solar energy.  Learning from their hard experience, we opted to have spare panels available just in case of similar circumstances.  We made a similar decision to have a spare battery on hand.

How Many Panels?  The decision to go with the Solar Word 275 watt panels filled in several of the parameters needed to determine the number of panels required to meet our electricity needs.  We already had established we needed to capture an estimated 7500 watts daily from solar energy.  The panels produce a maximum of 275 watts and they have an active solar cell area of 1.47 square meters for each panel (calculated from the number and size of solar cells in the panel from the spec sheet), and their efficiency for converting sunlight to electricity is 16% (from the solar panel spec sheet).  To complete the calculation, we need to figure out the amount of sunlight falling on each square meter of those panels during a day– the “insolation” values.

Getting the Insolation Values.  Insolation is essentially how much sunlight strikes the surface of 1 meter square.  If you want to know more about insolation try here for an entertaining read (trust me). There are several ways to verify how much sunlight is available at a particular site.  I used three different methods of calculation: one involving the average number of “Sun Hours” per day (aka Peak Sun Hours) from data available on line.  While this may be a fairly coarse approach, it provides a quick approximation.  You can try it here for your location.

The values we used for this calculation were:
  • Daily energy usage = 7500 watts, determined from a separate calculator, which indicated adding an additional 25% factor to allow for system expansion, and unsuspected system losses (such as dirt collecting on the panels during the dry season) or 6000 W X 1.25 = 7500 W
  • Days without Sun = 5 (AKA “days of autonomy” – note from Wilma - sounds like a good book title)
  • Lowest Battery Temperature = 60 degrees F (now that is nice! If we were back in Minnesota it would have been -30 degrees F).
  • Battery Bank Voltage = 48 (from wiring twelve 4 V batteries in series)

The output at this step is the number of watt.hours needed for the battery bank. 

They then go on to calculate the minimum number of solar panels needed to meet this wattage, and additional inputs are required and is redundant from an earlier calculator.  Sun hours for Belize (a measure of how much sunlight is received) = 4 (from the convenient insolation maps provided with the calculator).  It then calculates the total wattage needed from the solar panels.

Once you input the wattage for the solar panel you will use, it calculates the number of panels needed.  Our panels are rated 275 watts maximum under brilliant sunshine conditions.  This results out that we needed 12 panels to accommodate the estimated electrical usage.

When we discussed this with our Solar Installer, he advised adding an additional 25% for the number of panels, since that would improve the collection of solar energy on cloudy days, and minimize the depth of battery discharge, prolonging battery life. 

Have you noticed that an additional 25% is added a lot in solar estimates?  It was used to estimate the electrical usage (6000W X 1.25 = 7500 W).  It was recommended here to increase the number of panels from 12 to 15. It is also used to size the charge controller you need for charging the battery bank from the solar panels.  Furthermore, 25% is used in increasing the size the electrical wire conductors from the solar panels to handle the current – do you really want to know more? 


Fifteen panels it is!
OK, We Are Good to Go Based on Average Values -- 15 panels it is! However, the amount of sunlight varies on average from month to month.  Less sunlight reaches the panels during the rainy season, more during the dry season which is a bit variable, starting mid to late February and ending some time in June through August.  Less sunlight or insolation during the winter, more during the summer.  This is shown below using data from NASA on how much insolation is expected for a nearby site about 40 miles south of us (Punta Gorda, Belize, see the Magnetic Declination Map in the first installment for PG’s location relative to us) using data from the Table of SolarEnergy and Surface Meteorology.  We will be fine during the summer, but December and January may be more of a problem.

If you sum up the monthly values and calculate an average watt.hour per square meter per day (ignoring the different number of days per month) you get 4.69 kilowatts of sunlight per day hitting each square meter of surface on average.  But December's value is only 3.44 kilowatts, a 27% reduction from the average.

You can, as I also did, perform a de novo calculation on the amount of solar energy captured for any given month, based upon the insolation values given above per square meter of surface, the area of the solar cells in the panel (1.47 square meters calculated from the spec sheet for the panels), the number of panels (15) and the efficiency for converting sunlight to electricity (16% from the spec sheet for the panels).  For December, this calculates to a bit over 12 kilowatt hours on average per day, which is sufficient to charge the batteries and take care of normal electrical usage.  It is always good to verify the rough calculations from above are correct.  But if there are several days of uninterrupted gloom, we would have to use a generator for charging.

If you would prefer an easier online method, you can try this instead from the National Renewable Energy Laboratory in the USA, which I again found after doing everything above, which also has data available for international sites (see below).

After I put in the size of the solar panel array in kilowatts (4.125 kW, using the maximum wattage value, 275 watts times 15 panels), a few other parameters regarding the orientation of the solar panels (azimuth 180 degrees and 32 degrees of tilt – remember from the first solar blog?), I received the following calculation for our approximate location, which used data from a weather station 45 miles away in Puerto Barrios, Guatemala:

Yup, December is the only month that will be a little dicey, and will likely require some generator use after a series of cloudy days.  By the way, the 5,221 value at the top of the image and at bottom of the middle column is 5,221 kilowatts, or 5.221 megawatts of potential AC generated over a year that we can collect from a dinky little 4.1 kilowatt.hour solar panel system. Now we are the South Englishtown Electric Company!

Changing various values in the program could not improve the December monthly value by more than ~1%.  Changing panel orientation and or tilt was not effective.  So like I said in the first installment: don’t sweat the small stuff.  If we want more solar energy in December, we have to add more panels – it is good I have some spare panels on hand.  Alternatively use the generator a bit on cloudy days or be especially vigilant in minimizing power usage.

How does our 5221 Kw.hr/year compare with some other parts of the world?  Well here are some representative values:
Max. Kw.hr/year
Miami, FL, USA                  5975
Kona, HI, USA                    5386
Sydney, AU                        5861
Madrid, Spain                     5822
Berlin, Germany                 3670
London, UK                        3808
Torquay, UK                       4488
Kuala Lumpur, MY              5584 (using the closest weather station about 700                                                        miles away)

You can put in more locations into the program and play with this if you wish and it will use data from an International Airport weather station nearby (hopefully).  Intuitively, locations with fewer clouds have more solar energy available.  As you go north in Europe, the cloudiness increases and the insolation decreases (Madrid vs London for example). 

A bright spot however, does show up on the “English Riviera”, with Torquay having about 18% greater insolation than London.  This appears to have caused some hotel owners to install solar panels.  But it looks a tad overcast to me (LOL). 
Fawlty Towers Goes Solar!
Looks to be about a 10 kw.hr installation. Poor Mr. Basil! Did he hire Mr. O’Reilly on the cheap to install the solar panels? It would be better if the panels were rack-mounted and not sitting on the lawn, and not going around a curved driveway that changes the azimuth! (See re-runs of the television series acclaimed by the British Film Institute  “Best British television programme ever made” – Fawlty Towers !!!  (exclamation marks mine).  And thanks to Microsoft Office for their image manipulation programs used to create this chimeric image.

In the final installment, we will go over the easy, but nail-biting part for us: the installation of the system by the solar pro’s, and taking out some insurance on the solar energy system.


  1. Solar panels in England? Isn't that an oxymoron? Yours in Belize should be amazing. So glad you're making progress. Isn't your "Rule 1" of solar installation the same "Rule 1" for every single home project?

  2. You raise some good points, Mitchell! ;-) Rule 1 is universal and sometimes works in tandem with Rule 2 which states: "In exceptional cases, like yours, please allow for a 40% over-run on time and expenses."

    Seriously - we are extremely happy with the performance of our system.

  3. Yes, I think I can imagine Wilma playing the part of 'Sybil' in Fawlty Towers.{:))

    1. And here I thought I was shoo in for "Polly", Roy. :-( ;-)

  4. ¿QuĂ©?

    Manuel (aka Dennis)


Blog Readers -- your comments are invited. I would love to hear from you.