Monday, May 21, 2012

The Solarian Mirage

The sun shines bright
We live in an age of fear and guilt, because we realize that the oil and coal are running out, and we've (puportedly) carbonized the atmosphere to a level high enough to cause weather changes.

While the first point is true and quite serious, the second is debatable (I won't debate it here, not today at least). Green is the new gold, and people are apt to jump on the "Solar solves all problems!" bandwagon at the drop of a sun hat.

I don't buy into this meme, because unlike almost all of you, I actually lived along with my family on solar electricity for several years, and what I and my brother have forgotten about the practicalities of solar electricity is much more than most will ever learn.

Now though, I (obviously) live in an urban home, and last month I consumed 326 KWh of electricity which runs :
  • A fridge for 24 hours
  • TV for upto 8 hours
  • 3 CFL Lights for 6 hours
  • Desktop PC for 6 hours
  • 2 Laptop chargers for 12 to 16 hours
  • Water heater for 30 minutes
  • Microwave for 30 minutes
This is a very meager amount of energy consumption compared to most households.
Let's calculate what it would cost me to go solar, given that I need a paltry 10 KWh a day...

Generation
Let's assume I get 4 hours of peak sunlight every day, which is very optimistic, given that it is tropical weather and urban atmosphere to boot.

I would need get a power output of 2.5 KW from my panels. That DC voltage would need to be converted to AC, at an efficiency of about 80% (again very optimistic) so that makes it 3.125 KW.

Since I need power at night, I need to store energy in lead acid batteries, at an efficiency of 70% (not including the loss caused by charge control circuitry). This gives me 4.46 KW of peak power requirement from my panels.

I looked at some listings on e-bay for poly crystalline panels and they need about 5 to 7.5 sq. m. per KW of peak power. We will think about pricing later.

Thus I need anywhere from 22 to 35 sq. m. of panels, installed on the roof of my house, which is a lot!

Also consider that I would like to have about 50% extra power, so that In a span of two sunny days, I can save up enough energy for one cloudy day (once again rose colored glasses).

I wish I had a dollar fifty for every time I prayed for sunny weather back when I needed it.

Storage
As of now, there is no widely available alternative to Lead Acid batteries, and while the technology has improved in the 15 years since I last used them, they are still heavy, expensive and short-lived.
For my hypothetical green home, I would need to store power for about a week shall we say? Let's be more optimistic and say 3 days.

30 kilowatthours - that is 2500 Ampere Hours at 12V

The best deep cycle tubular batteries are about 200AH capacity so I need about 12 of them. At 75 KG a piece, that is pushing close to a ton of equipment, including a 5KW inverter and all the associated paraphernalia.

Cost?
About 15000 Rs a piece for the batteries
About 15000 Rs per sq. m. for the solar panels (based on rates of 300 USD /sq. m.  as seen on e-bay)
About 15000 Rs for a good 5KW inverter, I'm guessing.

Panel Life? They say 30 years, and it seems like that is a reasonable estimate.

Battery Life? Don't believe the manufacturers hype, if you can preserve 75% capacity of the batteries after 5 years with a few deep discharges in between, I will present you with an iridium asteroid.

So it costs me :
  • 4.46 * 1.5 * 15000 = 100,350 for the panels 
  • 15000 for the inverter
  • 15000 * 12  = 180,000 for the batteries
Total is about 2.8 lakh rupees initially, assuming I will do all the wiring and installation myself  (You could too,  if you just master one simple formula V=IR)

Every 5 years I would need to replenish the batteries, so I buy say 4 batteries to supplement the aging ones - That is 60,000 every 5 years or 12,000 per year after the first 5 years.

After 30 years... 
Assuming prices are constant, my power usage does not rise, sunlight remains plentiful and so on.

I would have spent 1500 a month on regular power from BESCOM - that's 1500 * 12 * 30 = 5.4 lakhs of rupees.

For solar, I would initially spend 2.8 lakhs and for 25 years after the first 5, I would spend a 1000 a month on battery cost = 1000 * 12 * 25 = 3 lakh rupees.

Does not really work out cost-wise!!

Let's look at what variables I could have misjudged by a great margin.
  • Panel costs  - perhaps they could be cheaper by another 15%
  • Battery costs - lead acids not likely to be cheaper than what I mentioned unless technology inevitably leaps and bounds, but halving the price of batteries would probably take a decade.
  • Sunlight - I counted ampere hours day after day in the Himalayas, where the skies are actually blue, and I declare that 4 hours of peak power a day is extremely optimistic, as is 2 sunny days for 1 cloudy one. 1 week without sunlight is not uncommon. And if you just deep discharge your batteries a few times, they will be scrap!
  • Rising costs of conventional power - this is very much what can throw my price numbers out the window
  • I don't consume much energy! I bet most households are consuming at least 2 to 3 times as much.
  • Energy costs of making the panels, the batteries, and the costs of installation (let me pretend I can't understand V=IR)
What about cooking?
That itself consumes almost as much energy equivalent as I use electricity. Once again there are only 2 people in my domicile, and we don't cook or eat much by most standards.
 
Then again, what about transport?
I burn 20 to 50 times more energy in the form of gasoline when I ride my bike - If I had a car, I would use probably 1.5 to 2 times as much of that again, not to mention the immense energy required to smelt, cast and forge three quarters of a tonne of steel and aluminum into an automobile!
Should I choose to use an electric bike and forgo the beloved noisy IC engine, I could reduce the energy needed by my bike to about a tenth but even then it's a very big amount.

Conclusion
So it all adds up to some dismal numbers, cost wise and carbon wise - and these numbers do not paint a good picture for solar. Don't even mention wind to me, because it is even worse.

I'm not saying solar is useless, but it's a bad choice, considering the demands of energy we have - and I haven't even considered heavy industry or jet transports or ships.

The wise thing would be to push for nuclear - whether you like it or not, in 25 years most of our energy needs will be from fission reactors. The French are doing it, so are the East Europeans - it works...

There's no other known way to run all those zillions of automobiles and air conditioners and what not. If you have huge amounts of nuclear energy, you can split water to make hydrogen and make almost totally emission free automobiles. It's been done decades ago. Even if hydrogen storage is an intractable problem, we can still use electric vehicles.

Nukes are the only technology that can be implemented right here, right now, to provide reasonably cheap and emission free energy for everything except air transport (and maybe that too, once they master hydrogen burning jets).

Anyway, 20 years from now, I will be reading this blog, either feeling like a complete idiot or a prophet! But somehow the idea of square kilometers of panels and tonnes and tonnes of batteries, just does not look very doable.