What a fluff piece. The author should run some numbers on the energy density and size of a battery that would be required to meet energy demand when the wind doesn't blow and the sun doesn't shine.
In this case where the grid has to fall back on non renewable sources of energy the utilities that own them are going to charge a very dear price to recoup their costs.
Just wishing for something doesn't make it come true. The battery technology needed is non-existent. The article says that batteries "will improve so much over the next two decades that homes won’t be dependent on the utility companies" isn't just wishful thinking. It's wishful thinking unsupported by known science. Similarly, in 1973 just as I entered college, nuclear fusion for power was also just around the corner. 40+ years later it's further away than ever.
Germany, one of the leaders in solar (even though they don't have the climate for it) is learning the hard way that just wishing for good outcomes doesn't make them magically happen. Here's [1] an article just from today that shows that they are burning more dirty coal than ever in their quest to be green.
Solar power doesn't work at night. In parts of Oregon and Washington it doesn't really work very well for about half the year, even in "daylight" (which is in quotes because it's often heavily overcast or raining here for much of that time).
Anyone who wants "unlimited" and "free" energy should get to work on figuring out how to store all that solar energy for when "the sun don't shine". Otherwise this article is about as useful as something pulled out from where "the sun don't shine"!
However, I'm not anti-solar, I'd love to see the deserts in Arizona, New Mexico, and Texas dotted with solar installations. I'd love to see HVDC used to transport all that energy to where it could be put to good use. But none of that infrastructure will ever be close to "free".
Ah. Venturebeat. Never shy about hyping the future.
The power industry was saying that nuclear power would be 'too cheap to meter' in the 1970s.
Ray Kurzweil does not make that mistake but is he right about the cost curve going forward? It's not like semiconductors where you get cost-reduction driven by a physical effect.
cost-reduction is also a function of productivity. We see that - at least since the industrial revolution - every doubling of production reduces cost 20-30%. It's the same whether we're talking about cars, shoes, planes, or solar panels.
First off, what's the decrease in cost if the production stays the same? If it's also 20-30% then the doubling isn't really relevant, is it? (One example might be the cost of a 75W incandescent bulb, which was essentially a commodity over several decades, but where the doubling time would be long.)
Second, how does one compare a car from 1950 to a car from now? They are built to very different standards and expectations.
The average purchase price of a car in the US in 1950 was $1,500. That's an upper bound of the production cost. Should I really believe that a modern car costs no more than $1,500 * 0.8^3 = $750 in 1950 dollars to produce? That's under $8,000 in 2014 dollars, while the average sales price for a modern car is more than $30K. I don't believe the profit margin in car manufacturing is anywhere near that high.
Can you explain how cars fit your description? Since my quick analysis says it's not all that meaningful to the discussion.
Also, the production cost of petroleum has not followed the trend that you suggest is universal.
Look at the capability & price of LED lighting over the last 20 years as their production has skyrocketed.
As you said, cars have simultaneously become much more capable (safety, reliability, fuel-effiencey, performance) while they have gotten cheaper. Your $1500 1950 car cannot be compared to a median modern car. Even the Mercedes S class from then 50s (arguably the most advanced car) cannot be compared to a Toyota corolla today, the corolla is WAY more capable.
As the production of wealth like 'air-bags', 'cruise-control', & 'direct-injection' doubles, they become cheaper and cheaper, even if the overall production of cars doesn't necessarily double at the same rate.
My statement of "reduce costs 20-30%" can manifest in greater capabilities as well as a reduced price. Both a 1985 and 2015 television could easily cost $1000 (inflation adjusted), but the 2015 TV would probably be close to 10x bigger, use 10x less electricity (OLED), and have 10x more overall capability. This is not only a function of global TV production, but all kinds of cross-pollinating technological productivity gains, from semiconductors to logistics. It's all connected so it's hard to make generalizations - and there are always exceptions - but economists have found that the "doubling production = 20-30% reduced costs" rule is reasonably starting point.
I rather strongly dislike the use of "20-30%" for something as inherently subjective as measuring the differences between a 1950 and 2014 automobile. How many percentage is an airbag? How much better are seat belts? Where is a study which has tried to quantify this?
The same holds for LED lighting. We aren't producing the same lights as 20 years, so it's a function of far more than productivity.
I believe you are talking about the experience-effect curve, described at http://en.wikipedia.org/wiki/Experience_curve_effects . "This relationship was probably first quantified in 1936 at Wright-Patterson Air Force Base in the United States,[1] where it was determined that every time total aircraft production doubled, the required labour time decreased by 10 to 15 percent."
That curve describes about units in a production run, not aggregated over the industry across multiple types of units.
> all the technologies we study here behave roughly similarly: Information technologies closely follow patterns of improvement originally postulated by Wright for airplanes [3]–[8], and technologies such as beer production or offshore gas pipelines follow Moore's law [9], [10], but with a slower rate of improvement [8], [11]–[15].
The cost reduction in the chemistry industry ranges from 1% to 48% if I read the supporting information correctly. It also covers other industries.
Interestingly, it comments that:
> One automobile may be faster, while another is less expensive. For this study, we focus on one common measure of performance: the inflation-adjusted cost of one “unit”. This metric is suitable in that it can be used to describe many different technologies. However, the nature of a unit may change over time. ... We nonetheless use the changes in the unit cost as our measure of progress, in order to compare competing models using a sizable dataset.
but doesn't actually measure the automobile industry. Then again, according to my estimate, under that unit-based metric automobiles have a negative curve.
The paper continues:
> The crudeness of this approach only increases the difficulty of forecasting and makes it particularly surprising that we nonetheless observe common trends.
It might be because it's mostly talking about commodity production. The exceptions, like MonochromeTelevision, are under "Other Industries" and date from WW2 and the post-war era.
Photovoltaics, btw, is in the 20-30% range you mentioned. I'm much less convinced about cars and shoes.
Also, the Wikipedia page points out that could really just be seen as a an economy of scale. For example, Goddard C (1982) Debunking the learning curve. IEEE Transactions on Components, Hybrids, and Manufacturing Technology 5: 328–335. doi: 10.1109/tchmt.1982.1136009
That PLOSOne article tries to test that, coming to the tentative conclusion that improved productivity is a better for short time scales, and Goddard's model for economy of scale is fine for long time scales. This suggest to me that economy of scale and not improved productivity may be the more relevant factor for the decade long trends you're talking about.
In this case where the grid has to fall back on non renewable sources of energy the utilities that own them are going to charge a very dear price to recoup their costs.