[jarrettc]

The article suggests that solar panels could be an "existential threat" to power companies. For those of you who are familiar with the math of solar panels, how realistic is that prediction? How much of a normal home's energy needs could be supplied by solar panels, assuming the panels covered the entire surface area of the roof? How much is the ROI on solar panels likely to improve in the foreseeable future? (I've heard estimates of 5-8% annual return for 2014.)

[djrogers]

I installed a 10kw system ~15 months ago on my 3600sq ft house. Average electricity costs was ~$6k/year before the panels, the first year on solar to total cost for the year was $300.

That $300 worth of electricity for the year is before I've been able to talk the wife in to replacing our incandescent bulbs, swapped out our old pool pump, etc etc - I could easily get that down to below 0 with a small investment in bulbs. So in my case at least, it was trivially easy to get to a net-zero usage.

A few $$ notes - the system cost $37k, I got 11.1k of that in a tax credit from the feds, so my net cost was 26k. I'm saving 5.7k/yr, which means my break even point for the system is 5 years. In 5 freaking years I'm making money from my solar panels! And I don't have to write a $500 check to PG&E every month. It's a beautiful thing...

BTW - the above numbers (plus my severe aversion to debt) are why I'm so against solar leases. If we had a solar lease, we'd still be stuck with monthly payments, would have trouble selling our house, and would be stuck for 20 years.

[pkaye]

Where do you live? Do you use electricity for heating? I'm in California and for a 1500 sf house, it costs us roughly $600 in electricity and $600 is gas per year. This actually an old house with some insulation. I think it would be more efficient for me to improve the insulation and water heater/furnace before even considering solar panels. I wish they would subsidize these other improvements as much as solar panels.

[djrogers]

I'm in an unfortunate area of California where my tier usage is tied to parts of the bay area, but the weather is closer to the central valley.

Our AC and pool pump are big parts of that $6k/year usage, and while AC is a luxury, when it's 105* for a week or two straight and you've got little kids around, it sure is nice.

There are also 2 other factors that feed in to our electrical costs - I work from home, and we have 2 little kids. That means our home doesn't get to shut down during the day. AC, fridge door opening, lights, etc., it all adds up when 4 people live in a house 24/7 instead of being gone during the hottest parts of the day.

[ensignavenger]

You mention the AC and pool pump- and then state that the AC is a luxury :)

[danans]

California does subsidize upgrades to gas and water heaters and insulation. Check out https://www.energyupgradeca.org/en/save-energy/home/take-con...

I'm in the bay area just replaced my furnace, water heater, and insulated ducts (an 11k job), and I'm expecting a $2800 rebate from the BayREN program (the local program that implements Energy Upgrade California).

https://www.bayareaenergyupgrade.org/program-overview

[Roboprog]

Like "djrogers", I also live under (let's name names) PG&E and their "infernal" tiered pricing. Try paying 30 to 40 cents a kW/h, and also working from home with kids running around during the summer. I live in the Sacramento area, but outside of "SMUD" coverage, so I have to have PG&E. When it is about 100 degrees +/- for 3 months solid, A/C is "life support", not a convenience.

$400/month electric bill, very typical for summer.

[maxerickson]

Have you investigated the insulation/sealing status of your house?

[fnordfnordfnord]

he said $.3 or $.4 per kW/h. It's not his insulation. That's super expensive.

[maxerickson]

A high price of power would magnify the effects of poor insulation.

Someone paying $5,000 a year for electric has strong incentives to make improvements, so I was wondering about the specifics, not trying to make a brilliant suggestion that all they need to do is put up some pink foam.

[toomuchtodo]

I'm absolutely shocked at the PG&E pricing. I'm in Chicago buying power from a 3rd party provide delivered via ComEd, and I'm only paying 8-9 cents/KwH, and that's more than from ComEd directly so I can buy wind power exclusively.

[secabeen]

Yep. That's the standard power pricing throughout California. SCE and PG&E are trying to flatten the tiers somewhat, which will reduce costs for high-end users, but a $300/mo bill is not uncommon for a large house.

[monksy]

Is it integresys? If so I opted out of that and saved quite a bit of money. (It's the fee that gets you) Going direct with ComEd has it's advantages. Also Chicago has a lot cheaper power than other areas.

[toomuchtodo]

FirstEnergy. There was no additional fee in Schaumburg to use them, and I pay 6.65 cents/KwH for 100% wind power (until 2017, at which point I expect the price to plummet).

Northern Illinois has extremely cheap power due to Exelon's nuclear generation capacity (Ameren in downstate us primarily coal-fired). I expect the price in IL from ComEd direct to start going up, as they're going to use their smart meter rollout to start pushing time of day metering (as they should) vs flat rate per kwh pricing.

[Roboprog]

Well, PG&E pricing starts at around 10 or 11 cents KwH, but once you plug in a refrigerator (or so), that tier is pretty much used up, and you start sliding up the scale.

[dntrkv]

I grew up in Sacramento and that's completely absurd. Our electric bills topped out at around 200, with 7 people, living in a ~3,000 ft^2 house, and someone was always home running the A/C. $400/month is insane.

[Roboprog]

Another thing - in Sacramento county, you probably had SMUD, but Placer, Yolo and El Dorado have PG&E, like most of the rest of Northern California.

[seanflyon]

How long ago did you electric bills top out at $200?

[dntrkv]

Good point, that was a while ago. 2008 was the last year I lived in Sac, so around then.

[maratd]

I'm in NJ and my numbers are the same. I think it's because we use gas for heating/cooking/hot water. If everything was electric, I can imagine things costing almost twice as much, if not more.

[hga]

Where do you live?

Where I live, in SW Missouri, capital and maintenance costs might be higher due to weather (e.g. high winds, hail), and it would be useful during the summer---which the electric company would be generally thankful for, less $$$ peaking power needed---but much less so in fall and winter when it's frequently overcast.

Also, if you live in a place with high electricity rates, like California, the case elsewhere is less compelling. I probably pay on the order of 1/2 of your savings for the total power to handle a building that's almost certainly bigger and thirstier than your's (numbers on request).

[mrfusion]

$6K/year sounds really high no? My neighbor pays $100-150/month for a 3200 sqft house.

[Shivetya]

Welcome to a fully regulated rate structure, throw in the combined weight of years and years of "its for the environment" and the costs go to the people who have the least voice and for many the least ability to pay.

The have so regulated pricing and punitive structures into service there that they have to have assistance programs just to alleviate the burden on middle income and lower consumers.

[Igglyboo]

He says he has a pool and if he lives in california then that pump is probably running all year, probably adds a few hundred a month to his bill.

[jere]

Yea a $500 per month electricity bill sounds steep to me. When my bill hit $400 in the winter, I realized I had a problem with my water heater.

[HeyLaughingBoy]

That's what I thought. My house is about the same size as his and completely exposed to weather. While I live in not-so-balmy Minnesota, it does get well above 90F and humid regularly in summer. Not to mention that the only non-electric heating appliance is the furnace.

I get irritable if my bill hits $250, I'd be freaking out at an average of $500/month.

[ryanhuff]

How much power do you consume a month? Pool heaters can really kill your electric bill.

[chdir]

How many units does $500 fetch where you live? Would be curious to know a breakdown of the bigger portion of the pie in those units.

[baddox]

How long are your panels expected to last, and how much maintenance do they require? Is the 5 year break even point pretty much guaranteed, and considering those things?

[djrogers]

They've got 20yr warranties, and require zero maintenance where I live. If you have more rain/snow than I get, or perhaps salt spray etc, you may need to do some cleaning.

Once every few months the monitoring doohickey needs a reboot, but the panels work with or without that.

For my personal situation, unless PG&E reduces their power rates over the next 5 years, I'll absolutely hit that break-even point.

Panel production will degrade over time - the warranty covers something like 3% the first year, and 1% every year after that. So at 5 years I will have less than an 8% reduction in output from the panels. That 8% will be made up by that time with LED bulbs, a variable speed pool pump, etc. Heck, right now out of the ~30 ceiling cans in my house, only my office (4) and front porch (3) are LED. The rest are full 60W sucking incandescent bulbs. The chandeliers are also 60W bulbs too, accounting for another dozen or so lights that are on quite frequently.

What can I say - my wife hates CFLs and barely tolerates LEDs.

[baddox]

You mentioned the reduction in your energy bill, but do you happen to track how much of that comes from selling back to the grid and how much comes from you using your solar power directly? If you're generating 10kW a few hours each day, I would guess that's the only time you're selling back. Would there be any point in an energy storage system for you? I've always had this fanciful notion of having a solar array and building an underground flywheel for storage, but only because I'm riveted by flywheels.

I haven't looked into the ROI of CFLs or LEDs vs. incandescents in a while. I imagine they have improved. Lately, I've even seen some LED bulbs in stores that don't look horrible. Of course, I would buy them anyway just to save time changing bulbs!

[secabeen]

I have a similar system, and I sell back for about 80% of daylight hours. That credit covers all of my other needs in the summer (and then some), and about half of my usage during the winter months. When you originally design the system, you use current rates and policies to create a system that generates at most a net zero usage.

[fixedd]

I swapped all of the incandescent bulbs that run for more than ~ 15min/day out in our house for some 80+ CRI "warm white" LEDs I got off Amazon. My wife didn't even know I'd changed them till I told her. I noticed the difference for the first day or so, but I got used to it really quickly. I wouldn't hesitate to do it again, and any bulb that fails in my house will be getting replaced with an LED from now on.

Maybe you could slowly transition... replace one bulb a week over the span of a year and she may not even notice :)

[g8oz]

Wholesale panels are $500 a kW. That's a lot of markup you had to pay. Definitely much room for improvement on margins.

[secabeen]

Labor is a big cost, as well as permitting and the like. It's certainly a business with a lot of profit in it now. As long as the marginal rates of power are high, it's still break-even for the customer fast enough that they don't mind. The feds and the utilities pay the costs.

[Pxtl]

Where you located? Are we talking Texas or Michigan?

[crdoconnor]

Hawaii?

[mikeyouse]

A big portion of the existential threat to power companies is the possibility of price-spirals.

The cost of maintaining all of the power infrastructure for a region is relatively fixed, especially if people don't disconnect completely when they add renewables. This cost is generally recouped by a small surcharge to every kWh that's provided to customers. The possibility of a spiral arises when customers start purchasing significantly less power from the utility, which means the per-kWh cost for the remaining customers will have to increase. As this cost increases, the ROI of adding solar panels improves, so more people add solar. As more people add solar, the utility must increase prices to the remaining customers.. etc. etc.

There's also pressure on this spiral from solar installers since the 'soft costs' make up the majority of new installations. As more people purchase systems, the per-installation price will decrease due to installers getting bulk discounts for materials, efficiencies (physical and bureaucratic - dealing with permits, etc.) from previous experience, and a bigger base to spread out the costs of their own capital equipment.

Detractors think that utilities are fear-mongering to increase their operating and price flexibility, which would likely result in higher profits in the near-term. I don't think I buy this theory though, looking at my last bill of about $50, $28 was generation and about $25 was transmission & distribution (T&Ds). PG&E bills T&Ds based on system costs / proportional usage, so if I cut my power, my T&Ds would drop to almost 0, increasing the cost on everyone else.

Here's a pretty balanced article about the issue;

http://www.renewableenergyworld.com/rea/news/article/2014/04...

[AnthonyMouse]

It seems like the obvious solution is to make transmission and distribution a flat amount rather than trying to amortize it per KWh.

[mikeyouse]

It will likely come down to that but since utilities are typically government-granted monopolies, it will take legislative action to allow changes. It will be a tough political sell since it would actively discourage new investment in distributed renewables and socialize the costs so that the people who used the most energy (and therefore drive most of the demand for T&D) will pay the smallest proportion of their upkeep.

[derekp7]

So if the power companies are worried about people buying less electricity, why do they constantly send out power saving tips, and push things like CF and LED bulbs, etc? Is that only because they are required to by law?

[mikeyouse]

The EPA and various utility commissions mandate efficiency programs, and efficiency improvements are often tied to promises of allowed rate hikes, but utilities also want to shave demand where possible.

An ideal world for a utility would be one where everyone used a moderate amount of energy at all times. The reality though, is that peak consumption is typically 100% higher than the night-time lows. An example from a utility;

http://i.imgur.com/gPR7DNk.png

Most of the efficiency programs are targeted at lowering that daytime peak. You would need less overall generation (saving money on building new power plants) and have a much more stable system if you could flatten that whole curve.

[toomuchtodo]

Its cheaper to reduce consumption than build additional power plants. The cheapest watt is the one you never needed to generate.

[mikeash]

Both increasing and decreasing capacity is expensive. If people use more power, you need to build more power plants, and that costs a lot of money and is politically difficult. If people use less power, you now have excess capital investments in the form of power plants you built in the past.

Their ideal is for power usage to remain exactly constant, or at least grow at a constant pace that matches local realities. The population and their power use wants to grow faster than that. Efficiency helps to compensate for that.

You'll also note that a lot of the efficiency tips are aimed at peak consumption. That's not the case for light bulbs, but there's a lot of stuff around more efficient air conditioning. Peak power is much more expensive than average, and in many places the power company can't charge residential users accordingly. Air conditioners tend to be used at peak usage times, so decreasing that peak usage can save them a lot of money.

[blacksmith_tb]

I have a 4.8kW system on my 1700sqft home here in rainy Portland OR and I am net-zero (and that's with everything electric - heating and cooling, water heater, cooking, etc.). But I am also in a certified Passive House, which makes it quite a bit easier...

[bjelkeman-again]

Building passive houses makes so much sense. Did you build it yourself or buy it?

[blacksmith_tb]

Bought - I was very, very lucky to have gotten it. But the builder/architect who made mine have built six so far in Portland, none of which have sold for much more than market prices. There's a lot of debate about how necessary going all the way to PH is, there's a "Pretty Good" house movement of just doing more generous insulation / other efficiency improvements, which is also much better than doing the bare minimum required by code. Less is more, either way.

[abakker]

Without directly answering the question of total capacity, solar does have a few problems that are unlikely to ever make it an "existential threat" to commercial power generation, especially when done at home.

1. It is highly dependent on the amount of sunlight an area gets. This might seem obvious, but is also problematic in that it is difficult to build the correct capacity for climates that vary significantly over seasons. Solar is well and good in Los Angeles, since the weather is very consistent, and the amount of sunlight doesn't vary hugely over the year. In Seattle though, if you were to install enough capacity to be useful in the winter, you would drastically over-produce in the summer. Over production of solar is currently a problem, and can have significant detrimental impact on the overall power grid. Typically, when it drastically over-produces, it can cause blackouts. This is amplified by the fact that solar produces its maximum amount of output at a time when people don't consume the most electricity. This may eventually be mitigated by better batteries, and alternative power storage systems like Vanadium Flow systems.

2. Solar has a maximum capacity, and will likely always need to be mixed with other "on-demand" power generation systems. At present, these are Coal or NatGas generation stations that can be ramped up to meet spikes in demand. Effectively, these plants allow us to store energy chemically in the gas/coal and burn it on demand. very few renewable energy sources have the ability to be ramped up to greater production over short period of time. Again, this problem could be mitigated by substantial battery installations, where a small amount of excess generation from solar could charge batteries that could be drawn on to deal with spikes in demand.

3. Power transmission is always going to be a problem. Solar is less and less viable the further toward the poles you go. Though we could generate power in the southwest, there are practical limits to how far power can be transmitted over existing lines without too much loss.

In all, though it would be possible for Solar to operate a given house or building with on-site generation. It is very unlikely to threaten the power companies, which give us a low-cost, reliable, and simple solution that is capable of meeting our needs.

[djrogers]

<quote> Typically, when it drastically over-produces, it can cause blackouts. This is amplified by the fact that solar produces its maximum amount of output at a time when people don't consume the most electricity.</quote>

This is a pretty regionalistic viewpoint - in many parts of the world the peak solar production and peak electrical usage are highly correlated due to air conditioning use.

[abakker]

That is very true - I should have been more specific. Solar doesn't generate in the evening hours, when everyone is still consuming their AC, their Lights, their TV and many home appliances.

Pumping workloads (AC, Refrigeration) are fairly static loads on the grid though and don't account for very much (~7%) of the overall energy consumption (http://www.eia.gov/consumption/residential/).

The trouble with solar only really happens when their are spikes, either in over production, or over consumption.

[malyk]

This is where on-site battery storage "fixes" the issue.

Most people are away from home during the day while their battery charges and then come home for a few hours at night when the battery powers the house.

Now, there are certainly cases, perhaps even many of them, where it wouldn't really work out. But for me it's the ideal setup. My wife and I live in a 1300ft2 house and use 250ish kwh a month in electricity including recharging an electric car a couple times a week. A 1.6 kwh system would cover almost all of our energy needs. Couple that to battery storage so we could use the energy we generate during the day when we are at home at night? Perfect!

[huxley]

Is there any region of the world that uses more energy during non-daylight hours? Everywhere I've lived, peak demand is daytime.

[kiiski]

Where I'm at, today the sunrise was at 09:10, and sunset at 14:30. Peak energy usage is in the evening.

[abakker]

You'll see many power companies that offer incentives like this - https://www.pacificpower.net/ya/po/otou/ooh.html

it is still daytime, but typically not peak solar hours.

[gohrt]

Is it not possible to "turn off" a solar panels? Worst case, you have to throw a tarp over some of your panels for summertime.

[abakker]

It is certainly possible. The challenge when doing this is that the solar installation that you put in to be able to generate during the winter months ends up being MUCH bigger, and more costly. Most of the ROI models on solar assume that when you do this, you'll be able to "sell" your excess capacity back to the grid.

Many/Most systems I have seen in southern California contribute energy to the grid directly, and power is still purchased through the grid like normal. When your panel is generating, the power you generate and consume balance out and your power is "free", but you don't consume power straight from your panel (a flawed representation of how electric flows, I know).

The challenge with these implementations are that the grid itself handles the power generated by panels, and can't turn them off. The reason they do the installations this way is so homeowners stay connected to the grid to even out their energy spikes, but also to save the homeowner the cost of battery storage systems. In this case, the grid acts as "storage" by accounting, rather than by real storage of energy.

[maratd]

This is easily solved if the utilities get properly involved.

My utility has been pushing a thermostat that they provide at no cost and will even give you a bit of cash for installing. The reason is that this thermostat is hooked up to the utility through the internet and can be shut off by them during a spike or an emergency.

I just got new smart electric meters installed too.

If those meters were a little bit smarter, connected over the net, and had the ability to restrict flow ... they could easily install devices on a per house basis that would protect the grid. That would put the ball back in the homeowners court and the homeowner can implement a solution to shut off the panels when the utility isn't interested in purchasing power back.

[marvin]

Just install an outdoor radiator (electric heater), perhaps with fans if outdoor temperature is very high, and burn off the excess electricity. Switch the radiator on or off by computer. Dirt simple and cheap.

[findjashua]

Re #2: is it be possible/feasible for panel owners to sell the extra electricity back to the grid?

[abakker]

This is precisely how most home solar installations work. They provide power directly to the grid and offset the power that they still draw from the meter. Most solar installations feed power to the grid at the meter, so the power company still knows how much power you consume, they just also see how much you provide and then only charge you the difference.

it is this setup in particular that is problematic for spikes in generation, since the power company can't just disconnect your solar array if it is adding too much power at a low demand time.

[gerbal]

I think it's that power companies depend on a very broad customer base to support their substantial capital investments. When you start removing a small portion of customers the cost of power to everyone else starts going up. And if those customers aren't consistently off-grid, i.e. they still need grid power sometimes, that can actually raise power costs.

Increasing generation capacity to meet demand spikes is expensive, and requires having large, expensive power stations waiting at the ready, just costing money.

[nobodysfool]

Well, one of my friends in Arizona has not necessarily his whole roof covered, but he is able to generate about 90% of his yearly energy usage. The only problem is that he doesn't have batteries installed, so he's selling his electricity to the local power company and buying it back. With batteries and a slightly larger system, he'd meet all his energy needs himself. It's definitely possible. One thing that helps is the hot water heater is solar powered as well.

[mrbabbage]

This doesn't answer your question about residential solar, but it's a useful primer on economic analysis for utility-scale operations.

http://www.eia.gov/forecasts/aeo/electricity_generation.cfm

The interesting bit here is how wind is already cheaper than any other option save combined cycle natural gas, and wind is on track to surpass natural gas.

[kefka]

I've been interested in the viability of those wind towers. The ground is blanketed with plastic tarp about 10ft above ground. It slowly tapers to vertical (approximates y=10^x curve). In the vertical section, there is a scaffolding with wind turbines.

The idea is that the sun heats up the ground. When the wind starts moving, it goes up the chute, spinning the blades.

It was theorized that a 1mile diameter ground cover + tower would give 500 MW of power.

[ilyanep]

I'm making a lot of assumptions here, as I'm not an expert on solar panels (took a class that touched on them once, but I'm mostly just Googling here):

* Typical efficiency of a consumer solar panel to day is about 12%

* Average of 6 hours of sunlight a day over a year, sunlight providing 120 W/m^2 [1]

* Rooftop of about 80 m^2

Total energy over a year comes out to about 2500 kWh.[2] Apparently the average household energy usage in the US is about 10000 kWh[3], which would imply that you could get about a quarter of your energy from solar.

A few caveats:

* Apparently some places get up to 2200 kWh/m^2 per year of sunlight[4], which would bring your solar panel total to about 21000 kWh a year.

* Not entirely sure how big your average rooftop is.

* Solar panels have been created with efficiency in the 40s of percent[5]

Hopefully I'm not completely off with some of these, but it seems reasonable that a rooftop system can provide most or all of your power in the near future, depending on how sunny it is where you live. If it's not sunny, you might have a lot more problems, though.

[1] According to https://en.wikipedia.org/wiki/Sunlight, the World Meteorological Organization defines sunshine as a state of receiving at least 120 W/m^2

[2] http://www.wolframalpha.com/input/?i=365+days%2Fyear+*+6+hou...

[3] http://www.eia.gov/tools/faqs/faq.cfm?id=97&t=3

[4] http://www.nrel.gov/gis/images/map_pv_national_lo-res.jpg

[5] http://www.sciencedaily.com/releases/2013/09/130923204214.ht...

[maxerickson]

Power is quite a lot higher:

http://en.wikipedia.org/wiki/Insolation#Earth.27s_insolation

(the 250 W/m^2 there is the 24 hour average)

[breischl]

From the link

>> Ignoring clouds, the daily average irradiance for the Earth is approximately 250 W/m2

Emphasis added. Seems like clouds might be a non-trivial factor, yes?

[maxerickson]

Yeah, clouds matter (obviously, locations vary). GP used an extremely conservative value for the energy available, I was just pointing to a less conservative value.

(The difference between the values is ~8x, I doubt clouds have that big an impact in very many locations)

[clayton21]

To further comment on djrodgers:

His ROI in 5 years but can possibly be less. Depending on the state/market/utility, the owner qualifies for local rebate incentive, homeowner solar water heating rebates, local labor substantability rebates and SRECs. SRECS are energy credits that are similar to forwards contracts that can be traded on an exchange. The federal tax credit (ITC) is 30% of the purchase price but expires in 2017 to 10%. But, forecasted solar cost drop to 2017 would far cover the itc.

[abakker]

I would point out that most ROI models for solar look so good because of very generous subsidies. If subsidies aren't counted, the ROI time is not nearly so good.

Additionally, the Total Cost of Ownership of Solar will be higher in areas where panels/solar systems are prone to damage from weather etc. Wind, Rain, Snow, freeze/thaw, dust can all damage solar arrays, or negatively impact their performance.

If there are batteries involved, those will likely also need to be replaced before the system has initially paid itself off. Even if batteries get cheaper and better every year, the disposal fees for old batteries and their current inability to deal with house-sized loads well can easily add significant cost.

I don't want this to sound like I am anti-solar. I am a big proponent of solar, but there are many problems with the current system. It is a somewhat unfair comparison no matter what, since the energy we buy from fossil fuel generation does not accurately reflect the cost of the pollution it makes or the environmental harm it does.

[djrogers]

"I would point out that most ROI models for solar look so good because of very generous subsidies. If subsidies aren't counted, the ROI time is not nearly so good."

True, without the 11k rebate, my break-even point would have been 7 years instead of 5. I'd still have done it...

[abakker]

Thats pretty good - You also have the benefit of getting alternative return on that 11K upfront, which helps.

Depending on whether you sell your house or not, I assume that the solar system will also generate a reasonably high ROI as an improvement to the home? It would be interesting to see what the real payback time period would be if you were to try to sell the house before its usage ROI were realized.

i.e. could you buy a house, add solar, get the rebate, sell the house after 2 years and still get a return on the solar investment money because you'd improved the value of the house?

[dorfsmay]

Kurzweil said their efficiency doubles every two years, and that they would be able to meet 100% of our energy needs by ~ 2030:

http://www.livescience.com/4824-solar-power-rule-20-years-fu...

[teraflop]

> Kurzweil said their efficiency doubles every two years

According to that article, he said the "use" of solar energy doubles every two years.

Efficiency has been gradually improving, but it's definitely not doubling every two years. Every few decades would be closer; that said, using an exponential curve to model a function that's capped at 100% doesn't really make much sense in the first place.

http://en.wikipedia.org/wiki/File:PVeff(rev141113).jpg

[pwnna]

I believe there was a report from US utility companies saying that solar will severely impact their revenue.

http://www.eei.org/ourissues/finance/Documents/disruptivecha...

[maxerickson]

There's a lot of parameters (load, location, roof size, season), but I think the answer is generally > 100%.

On the other hand, cities don't have the surface area to account for industrial consumption.