I don't know that this is what happened, but it seems likely.

1) Apple wants to use sapphire for main glass in Iphone.

2) Apple signs contract with GTAT to supply the sapphire, including a pile of money to build sapphire production facilities.

3) Apple pushes all risk onto GTAT. IOW, if Apple decides not to use the sapphire for the displays, GTAT has to repay the pile of money from step 2.

4) Apple does not use sapphire. GTAT can't repay money because they already spent it building sapphire production lines which no have no demand.

5) GTAT declares bankruptcy.

The reason it is difficult to handle the load swings is that coal and nuclear plants can not be quickly turned off and on. For example, it can take a full day or two for a coal power plant to go from cold start to full power output. We do have some power plants that can react quickly, natural gas can react in a few minutes with the newest plants to an hour with the older ones. Hydro is very fast and can be switched off and on in seconds.

  So take a typical grid that is getting half it's power from coal, maybe 10% from nuclear, and the rest from natural gas and hydro. Now do what Germany did and put in enough solar PV that 5% of the total energy of the year is coming from solar PV. As the previous article showed, the instantaneous peak that results from this is as high as 50% of the total demand. In Germany's case, when that happens, all the gas and hydro plants are off and all their power is coming from coal, nuclear, and solar. And there is the breaking point. If solar increases any more in Germany, when solar is peaking, they either have to start rejecting some of the solar power, the coal power, or the nuclear. The solar can not be rejected under current German law, which leaves only coal or nuclear. And those boys really can't turn off their plants very well. The result is that they push the spot price down to encourage anyone who can respond to the spot market to take power.

  Germany has already experienced negative electricity prices in the spot market because of this problem.

  Now imagine Germany doubles the amount of solar PV they have. This would result in them getting around 10% of their total electrical energy from solar PV. The instantaneous peaks would be as high as 100% of total demand. What is supposed to happen then? The obvious thing to do is to simply cap solar PV to an instantaneous peak of around 50% of total demand. That happens infrequently enough that Germany could still probably get 9.5% of their total energy from solar PV. That only would reject about 5% of the power the solar panels could have collected. That's reasonable technically, but illegal under current German law.

  If you look at this paper, http://papers.ssrn.com/sol3/pa...

  The authors introduce the concept of system-LCOE. LCOE is levelized Cost Of Energy. It is a common utility term. The idea is to sum up every single cost to produce electricity and reduce it to a single number. Historically, LCOE ignores any issues with integrating with the grid. System LCOE is meant to try to incorporate both LCOE and grid integration issues. If you look at their graph on page 19, you can see how the cost of integrating solar PV increases quite dramatically as the share of energy coming from solar increases. If their graph is correct, Germany is paying in the ballpark of 15% extra to integrate into the grid.

Here is an article specifically written to address the type of confusion you are experiencing.

http://www.iflscience.com/tech...

Germany is already past that.

http://www.ise.fraunhofer.de/e...

In 2013, Germany got 5.3% of their total electrical energy from Solar PV.

http://www.theguardian.com/env...

Germany hit over 50% of their instantaneous electrical power coming from solar on June 9th of this year.

  I think you are confusing instantaneous and total power.

I agree that has been historically true, but that's changing fast. Once Solar PV penetration gets to the point where about 5% of all the electricity is coming from solar PV, it starts to get really expensive to handle the load swings. To be clear, I mean 5% of total electrical demand for the year. That means the instantaneous peaks will be in the ballpark of 50% of grid energy coming from solar.

  Most countries are very far from this point, but Germany and Italy are there today. Both countries have dramatically slowed their adoption of solar PV, mostly because of grid integration issues. All of those issues would be solved with cheap storage.

I've consulted in big companies for quite a while. My experience has been that most of time, most people are trying to obey all the laws. That said, yes, "the line" does get crossed. In all the cases I've personally seen, "the line" was crossed either because of ignorance or for precisely the reason you state (the fine is lower than the expected profit).

And that's why this case is astonishing. Steve Jobs went so far over the line, he might have wound up in jail. That's something I've not seen. You know why no banker went to jail? I've seen this shit in meetings. Someone proposes something that is illegal. The discussion then focuses on costs and profits. It then moves to plausible deniability and the chance of going to jail. If the conclusion is that there is the slightest chance someone will go to jail, that's it. That idea is dead dead dead.

  Steve Jobs, like the Honey Badger, didn't care. He left a trail, IN WRITING, that could have put him in jail.

Insanely illegal.

Yes you would.

Fortunately, we already blanket a pretty significant portion of the earth with buildings, roads and parking lots. Put solar on all the buildings and cover the parking lots and you are well over half of the way there.

Here is the NREL report on this subject.

http://www.nrel.gov/docs/fy04o...

NREL states we would need 00.4% of all the land in the USA to go 100% solar electric. The report uses existing PV efficiencies. By the time we could possibly be near something like 100%, efficiencies will be higher and that land requirement will be down to something like 00.35% or lower.

Tesla's model S can already go around 270 miles on a charge. The next generation of batteries (in test cars right now) just about doubles that. How much range do you need?

The google self-driving car has already shown itself to be insanely good at avoiding crazy human drivers. Even going as far as swerving out of the way of human drivers trying to ram it. The only way autonomous cars will be a traffic hazard to human drivers is if the production cars take a HUGE step down from the existing prototypes. That's just not going to happen.

    I little bit of that is here http://www.technologyreview.com/news/520746/data-shows-googles-robot-cars-are-smoother-safer-drivers-than-you-or-i/

You cite factors that fall against solar, but miss all the ones that fall in solar's favor. The biggest is peak shaving. In many areas, usage peaks coincide with when the sun is shining. Peak power is the most expensive power. Imagine building a power plant and running it seven hours a year. Welcome to peaker plants. That's some hellishly expensive electricity. In places like Hawaii, Texas, Arizona, and southern California, when people put more solar PV in, the utility needs fewer peaker plants. This is HUGE. You know how much credit most utilities want to give to solar for that? Zero.

    But if the utility does something to eliminate the need for a peaker plant, you can bet your entire net worth the utility will be asking the rate commission for higher rates to reward them.

    The best work on this subject (trying to figure out what price has no one subsidizing any one) is coming out of the Rocky Mountain Institute. A good starting place is their survey of existing literature (http://www.rmi.org/Knowledge-Center%2FLibrary%2F2013-13_eLabDERCostValue). Austin electric also appears to have done really good work in establishing what they call a "fair value of solar". By their measure, the fair value of solar in Austin is currently higher than the retail rate. As more solar is added, this rate will fall. The rate is assessed annually.

> BTW, they are also paying about 35 cents per kilowatt hour,

    Most of that is taxes.

    Current German solar feed in tariffs for solar range from 11.8 to 17.5 Euro cents per KwH. The 11.8 cents (which is what larger installations get) is actually cheaper than new coal plants.

http://www.germanenergyblog.de/?page_id=14068

    If you want to see how an upper bound on how much renewables are really costing German consumers, look at the EEG surcharge. That funds all the feed in tariffs paid to renewable producers. I say upper bound, because renewables get no credit for the downward pressure they put on conventional electricity prices (google merit order effect).

http://www.germanenergyblog.de/?tag=eeg-surcharge

So far this year, Germany's gotten about 5% of their electric production from solar and Italy is around 7%. Solar is way cheaper now than when those countries installed the bulk of their solar PV.

    So yeah, tripling will do very little. No reason not to hit up a factor of 30 to 50.

Yes. And it will almost certainly do it better than almost any human driver.