Is effective solar power further away than we had thought?

That is the topic of my latest Bloomberg column, here is one excerpt:

The first disquieting sign is that solar companies are spending only about 1 percent of their revenue on research and development, well below average for a potentially major industry. You might think that’s because things are going so great, but some major solar users may have already maxed out their technology. According to Sivaram’s estimates, four of the five most significant country users — Italy, Greece, Germany and Spain — have already seen solar energy flatten out in the range of 5 percent to 10 percent of total energy use. The fifth country, Japan, is only at 5 percent.


A common view is that solar power will come into its own once batteries and other storage technologies make steady improvements. Yet Sivaram notes that lithium-ion batteries in particular are not well-designed for storage across days, weeks and months. Also note that about 95 percent of global energy storage capacity is from hydroelectric power, a discouraging sign for the notion that solar energy storage is on a satisfactory track.


Solar energy has great potential for emerging economies, but some very basic preconditions are not in place. India, for instance, would need to end its kerosene and electricity subsidies. Freer trade in solar technologies is found in Tanzania and Rwanda but not always in West Africa.

My column draws heavily on Varun Sivaram’s forthcoming Taming the Sun: Innovations to Harness Solar Energy and Power the Planet, Amazon link here.  This book is full of useful information, a pleasure to read, and more generally a model for how to write about science, technology, and policy.  It will definitely make my 2018 “best books of the year” list.


Solar energy has great potential for emerging economies, but some very basic preconditions are not in place.

Like the basic technology required for it to work? The batteries are not remotely suitable - and somehow I don't think it will be India that will be perfecting them.

The delusions about solar power are remarkable. More so because the problems are obvious and have been obvious for a long time. Willy Leys' Engineers' Dreams pointed out that solar power was great but was not practical without some better form of storage.

The present boom seems to be mainly a by-product of computer chip manufacturing. So not really suitable for Africa. As long as well meaning people to continue to sell Africa naive dreams, they will remain poor. The only viable technology that could help Africa is coal and lots of it. Presumably Africa has lots of it.

'The batteries are not remotely suitable'

You mean that charging a cell phone battery from a solar panel won't work? Really? Especially as the comparison is completely apt - just as a cell phone network dispenses with the expensive infrastructure involved in transmitting electrical signals over wires over extended distances, a solar panel used to charge cell phone or lantern batteries also allows one to dispense with the expensive infrastructure involved in transmitting electrical power over extended distances.

The problem is the duration of storage. Everyone expects you just need to store the energy for a day or so (right?). But in reality, for a near-100% renewables grid, you need to store it for months, not hours. The system is dominated by long-cycle variation, not short, and that controls how long your storage has to work for. There's significant battery degradation and energy loss on those timescales. It's hard for non-specialists to understand just how much storage would be needed in near-100% renewable solutions.

Take the example of the southern UK (not the best place for Solar power, but it's not too removed from European insolation averages and distributions). This receives ~75% of it's total solar energy over "summer" and ~25% in "winter". So if you built a Solar PV grid _just_ big enough to supply the country's electricity needs (ha!) then you would have essentially 2 choices.

1 - Build 3x more than the amount of panels that you would need (on average) in summer AND enough storage for diurnal variation to handle the winter peak demand (actually, you would have to build a bit more than 3x, but that's the ballpark). For the UK, this requires a total average (NOT baseplate) solar PV capacity of 160GW and about 1000 Gwh of storage. This is hilariously expensive.
2 - Build 1.5x more than the amount of panels you would need (on average) in summer. AND Build sufficient storage to carry 1/3rd of ALL the energy generated in summer over to winter, to balance the annual load. For the UK this is about 72000 Gwh. The storage capacity of the largest pumped hydro systems is about 10 GWh. Batteries cost about £100/KWh. This is hilariously expensive.

Of course, neither of these "solutions" handle bad weather years or even seasons, and can be ameliorated to some extent by tidal (equally expensive) or hydro (no more than 5% of total mix) and countervailing wind. But the raw numbers are so big it doesn't change the costs from "huge". It gets a LOT easier if you allow a substantial Nuclear baseload, or a substantial conventional component (~50%) of total supply. But the storage amounts and landscape required are truly awesome under any large scale renewable deployment.

I wanted to reconcile this good comment with mine below in which I state that long-term (multi-day) storage is irrelevant and not economically feasible.

I agree with Alastair that if the goal is 100% renewables with anywhere near current reliability, long-term storage would be crucial. If there is 100% RE, then there are no other options and so batteries would be the only place to look for electricity when the sun isn't shining.

However, getting from 5-10% up to 100% involved a lot of steps and, at the end of the day, 100% RE is a cool thought experiment, but not practical over the next few decades.

So, if the goal is to move solar up to 20% or even 50%, the initial application of batteries will be almost entirely one day storage. In a 20-50% RE environment (particularly in a gas-based system) problems such as multi-day weather issues would be backed up by traditional fuels.

I stand by my claim that within our lifetimes, there will not be a significant market (>5% of battery applications) that are multi-day. Even in off-grid systems the cost of multi-day storage would be so high that it would make sense to deal with weather problems by increasing tolerance for less reliable power and backing up any crucial operations with diesel.

It is also important to realise that the challenges of getting from 10% to 40% RE, for example, are very much different than those involved in getting from 40% to 100%.


Entirely agreed. At low overall RE penetration rates (<30%, say), storage cycles can be much smaller and driven by the diurnal cycle (and a ~2 week weather cycle, for North Atlantic nations). As you say; beyond a day or three it makes more sense to back-up with conventional forms because of utilisation effects. Conventional back-up starts to fail as an option once you move beyond 30-40% RE, and then you're stuck with horrible multi-month batteries. YMMV.

The excellent and free (on Piratebay) book by Alice J. Friedemann, "When Trucks Stop..." (2016) explains the battery problem for solar very well. In Table 17.1, for long-term storage of 80to 100% renewables, you need two to 63 days of storage in Germany and six to 30 days of storage in more sunny Europe/Med in general, which is not practical. That would mean 3 to 402 Tera Watt hours (TWhr) of batteries. To show how big that is, consider that the entire USA consumes 11 TWhr a day. So try storing almost 30 days of all the electricity burned in the USA every day...not going to happen anytime soon.

Friedemann's book is outstanding...the gold standard for energy books. Her prose is down to earth, and she does not talk over the reader's head, but she also cites research in copious footnotes, so it's not just sensational opinions. Friedemann points out that there are more practical ways to save energy, like using rail more (currently trucks are subsidized by passenger cars, and trains to move goods are not subsidized at all, despite being very energy efficient at moving goods).

Bonus trivia: what is the EROI (ratio of energy out to energy in) for solar? If EROI is zero, it means you burn as much energy as you produce (liquid coal is almost like this, and some processes for ethanol are also like this), which makes the technology a loser. For solar, estimates are as low as 6 (not that great, since extracting oil today is 15 to 30 EROI) to as high as 39 (fantastic! We should all convert to solar if this EROI number is true). Why the big range? To get the high of 39 EROI, you must assume fossil fuels are finite. Too technical to get into here, but that's how you get the high EROI. Needless to say, solar people advocate using the high number, which is a bit dishonest, since most of the time EROI is calculated assuming the input energy is anything you want, including cheap oil. Read my comments at TC's Bloomberg article comments section for more on this.


+1 for that.

The best accounting I've seen for Solar PV gives EROI between 1 and 8, depending mostly on climate and longevity assumptions. A lot of the early EROI calculations for solar seem to have grossly underestimated or neglected the installation and maintenance costs, and been very generous on half-life. They need to be updated. Obviously, you are not allowed oil to make them in the long term.

If we take the more optimistic numbers in a good climate we get a system with a slightly lower EROI than wind and slightly better than feudal agriculture. Which is ironic, because that's the future the Greens seem to want.

Don't forget that in the U.S. we pay some of the highest prices for solar panels. A German company with American manufacturing sued to get tariffs in place on Chinese solar panels, for alleged dumping. China had enormously built out there solar silicon and PV manufacturing before the market was there to support it, so many PV and solar polysilicon plants are shuttered and others are unprofitable. So what if they are selling below cost? I say we buy all we can at those low prices. Instead, government policy is to tax Chinese panels to support a higher price. The government is actually discouraging solar power!

'The problem is the duration of storage. Everyone expects you just need to store the energy for a day or so (right?).'

You realize that the explicit examples given were cellphones and lantern batteries, right? At least the batteries I use for such things last more than a day or so. It helps to buy batteries that retain 80% of their charge for a year, admittedly.

'But in reality, for a near-100% renewables grid'

And to think I explicitly said that charging such batteries removes the need for a grid for both telecommunications, and for light that is better than that provided by burning something like kerosene. There will never be a near-100% renewables grid in an industrial nation in the foreseeable (or possibly any) future. (Assuming that the hoary 40 year old idea of powersats turns out to be too hard to manage - though the idea of a space elevator removes a number of problems in terms of powersat construction costs and transmitting the power to the Earth's surface).

On the other hand, as an Australian commenter (Brak if I recall correctly) has written, his part of Australia seems to be moving that way, in part because it is more cost effective at the residential level. Talking about a rural village that currently has no connection to any grid at all being able to use solar power to provide telecommunications and lighting is a long way from talking about an electrical grid.

'But the storage amounts and landscape required are truly awesome under any large scale renewable deployment.'

Storage is its own subject (though one that is not completely insurmountable), but there is an immense amount of North Africa that would be perfectly suited for large scale solar use - like this currently existing project. 'Noor 1, the first phase of the Moroccan plant, has already surpassed expectations in terms of the amount of energy it has produced. It is an encouraging result in line with Morocco’s goal to reduce its fossil fuel bill by focusing on renewables while still meeting growing energy needs that are increasing by about 7% per year. Morocco’s stable government and economy has helped it secure funding: the European Union contributed 60% of the cost for the Ouarzazate project, for example.

The country plans to generate 14% of its energy from solar by 2020 and by adding other renewable sources like wind and water into the mix, it is aiming to produce 52% of its own energy by 2030.'

And do note the storage component - 'The plant keeps generating energy after sunset, when electricity demands peak. Some of the day’s energy is stored in reservoirs of superhot molten salts made of sodium and potassium nitrates, which keeps production going for up to three hours. In the next phase of the plant, production will continue for up to eight hours after sunset.'

For what it's worth, I've played with (frankly ridiculous) model solutions like powering Europe from a mix of a Atlantic wind and a huge PV farm in the Sahara.

The big problem with these (in addition to still very large storage costs to smooth out weather variations) is that grid and transport costs become significant part of total costs again. And even with a good ultra-high-voltage network, there are energy losses at those scales. It turns out electricity is one of the most expensive goods (by value) to move over any distance! It costs 40p to move that Saharan KWh to London!

It's physically possible in the way that sending a spaceship to Alpha Centauri is currently physically _possible_....

The Morocco facility is not PV. Neither is this one in California - 'The Ivanpah Solar Electric Generating System is a concentrated solar thermal plant in the Mojave Desert. It is located at the base of Clark Mountain in California, across the state line from Primm, Nevada. The plant has a gross capacity of 392 megawatts (MW). It deploys 173,500 heliostats, each with two mirrors focusing solar energy on boilers located on three centralized solar power towers. The first unit of the system was connected to the electrical grid in September 2013 for an initial synchronisation test. The facility formally opened on February 13, 2014.'

And not unexpectedly, the project took time to meet its goals - 'Contracted power-delivery performance of 640 GW·h/year from Units 1 and 3 and 336 GW·h from Unit 2 was met by 2017, following sharply reduced production in the first few years of operation, particularly in the start-up year of 2014.'

Why people remain fixated on PV escapes me - it has its place, and obviously the solar irradiation limit exists regardless of method, but for large scale projects, it is not exactly all that suitable. But plants such as Noor 1 or Ivanpah allow both a stretching out of useful generation time, and the ability to easily limit output if necessary.

'is that grid and transport costs become significant part of total costs again'

As the UK will likely be able to better quantify while importing power before EDF finally gets its facility up and running.

But in reality, for a near-100% renewables grid, you need to store it for months, not hours. The system is dominated by long-cycle variation, not short, and that controls how long your storage has to work for.

100% renewables? Suppose instead we aimed for a 50% or 75% renewables grid?

50% renewable for electricity production is probably a reasonable target over the next 50 year time frame, assuming a reasonable drop in energy storage costs. 75% is possible if energy storage costs drop at a higher than normal trend rate or if the electrification of passenger vehicles increases the size of the electrical generation market (reducing the size of the oil industry).

50% is probably reasonable. Iowa receives about 37% its power of its power from wind and other Great Plains states receive over 20%. And these states invariably have cheaper electricity rates than the national medians. At one time it was believed that there was a limit to how high a windmill could be. But the industry has seemed to have worked through many of the operational and logistical problems, There are lots of economies of scale in larger windmills and prices s have been diving.

I use solar panels to charge my phone when I'm in a camper on the top of a mountain. For work I have phones, computers, printers, lighted and conditioned space, and I generate economic activity.

I don't think people in Africa are on holidays all year round.

By the way this is the type of argument that solar activists throw around all the time. How much solar energy would be required to power a bake proofer/oven? One of my customers has one and bakes bread daily for their restaurant. Or how much solar panels would be required to run a Siemens board stuffer with oven? They use 20A 440v 3ph. Plus the power for lights, and the various equipment and space for preparing and handling the product in and out of this very productive piece of equipment? Another customer, a very small electronics manufacturing outfit.

The silly numbers quoted by solar activists are meaningless as they describe potential capacity under full sun, nothing about fulfilling actual loads over time.

What's a more unlikely miracle? An enormous breakthrough in the perenielly slow-progressing field of battery technology? Or reversing five decades of cost disease and rampant overregulation of nuclear power?

Getting solar to work will require enormous scientific and intellectual effort. But even given all that, getting public institutions to allow nuclear fission, may require even more genius.

Agree about the cost disease and rampant over-regulation. What were the estimates here again? At least 2/3rds of the baseplate cost was over-regulation?

So maybe in the West. But perhaps not in China. But then I expect the Greens to explain the discrepancy by saying that Chinese atoms are cheaper, or something.

The difference is that the Chinese happily poison themselves.

Amazing that you pair their safety record with a claim of 2/3 overregulation.

How many Chinese have died from nuclear power and its externalities? How many have died from coal and its externalities? (hint; a LOT. Those aren't clouds you see over Beijing)

Whatever else you can say about them, nukes are statistically the safest form of power per KWh. It's plain perverse to deny the empirical data.

I think you lost the thread. The proper comparison is to the US, where no one dies, but some energy forms come with huge costs.

Basically you nuke guys need to reject all real costs and claim theory going forward.

We solar guys use real costs, here and now.

"We solar guys use real costs, here and now."

You have yet to address any of Tyler's points.

I did below, but to restate, we can acknowledge current limitations without denying current opportunities.

Solar would not be experiencing "doublings," as it is now, without great opportunity.

Governments hosing money around are always a great opportunity. Whether anyone actually benefits is a different question.

"some energy forms come with huge costs."

It would be more accurate to say huge costs are imposed on some energy forms, while others receive huge subsidies.

And Hydro:

When, in 1975, about 30 dams in central China failed in short succession due to severe flooding, an estimated 230,000 people died. Include the toll from this single event, and fatalities from hydropower far exceed the number of deaths from all other energy sources. (more)

'and rampant overregulation of nuclear power'

It wasn't overregulation that caused this problem - 'When alloy tubing in one of the new steam generators at San Onofre leaked a small amount of radiation four years ago this week, engineers at Southern California Edison immediately instituted emergency protocols and shut down the nuclear plant.

Neither of the twin domed reactors on the north San Diego County coast have produced a spark of electricity since.

No one disputes what caused the failure — excessive wear in hundreds of tubes designed to drive hot steam through massive turbines is the confirmed culprit, numerous investigators and analysts found.

But what has become increasingly disputed since the plant went dark is the question of who is responsible for flawed replacement steam generators being installed and who should pay for the failure.

Edison, the San Onofre operator and majority owner, said it had no knowledge of design flaws that led to the Jan. 31, 2012, breakdown. Edison places the blame with Mitsubishi Heavy Industries, the Japanese firm hired to design and build the replacement steam generators.

“SCE was unaware of the steam generator defects until they were discovered after the tube leak in 2012,” spokeswoman Maureen Brown said in a statement. “It was up to MHI, as the designer and manufacturer, to decide what design features to include that would result in safe RSGs” or replacement steam generators.

Billions of dollars are at stake in the plant’s failure, and so far, the lion’s share of the tab is being covered by the ratepaying public.

Following a November 2014 vote by the California Public Utilities Commission, customers of Edison and minority owner San Diego Gas & Electric have been paying $3.3 billion of the $4.7 billion in identified closure costs, or 70 percent.'

And what a coincidence - the customers not receiving the benefits of a bright nuclear future are the one's paying the majority of the costs associated with a design flaw.

And of course, this is the sort of thing that leads some people to wonder if the nuclear industry is not regulated enough, particularly in light of how Southern California Edison lied, as seen regarding its claim concerning 'SCE was unaware of the steam generator defects until they were discovered after the tube leak in 2012.' As contrasted to this - 'Overlooked in the saga of the 2012 failure of the steam generators at the now shuttered San Onofre Nuclear Generating Station is what the owner told federal regulators about the equipment before it was installed.

Experts say a June 2006 meeting Southern California Edison had with the Nuclear Regulatory Commission is notable both for what the company claimed and omitted.

In the 19 months before that critical meeting with federal regulators, Edison executives knew of a flaw in the new steam generator design, according to records.

Steam too hot to handle, otherwise known as void fraction, was making its way through the equipment. A consequence could be excessive tube wear, the same problem that led to the permanent closure of San Onofre in 2013.'

The approval process for a nuclear plant design takes many years and requires many levels of regulatory approval.

And when it breaks, the approval process for a a fix to the nuclear plant design also takes many years and many levels of regulatory approval... and legions of lawsuits.

Meanwhile, the regulators involved are... being investigated for corruption.

If only there were laws against corruption...

My hope is that Canada will save us from ourselves by showing that performance-based nuclear regulation works. And that Terrestrial Energy will have a hit with their molten salt SMR in 5 years and start deploying them around the world. Maybe then the there will be enough motivation for the NRC to quickly adopt performance based regs for molten salt.

Molten salt is interesting. but I suspect corrosion is going to be worse than expected. They should get some years on pilot plants to prove it out.

"An enormous breakthrough in the perenielly slow-progressing field of battery technology?"

I'm not sure this is the correct framing. Battery technology gets more effective / cheaper at around 1-2% per year. You don't need any kind of enormous breakthrough to predict substantially cheaper batteries 50 years from now.

PV solar and wind power were never ready for use commercially. The first clue should have been the huge subsidies required. The entire PV and windppower industry was about extorting money from governments and tax payers. It was never viable and may never be viable commercially.

Depends on who "we" is. If we is me, then the answer is no. All these trends have been obvious. If "effective" solar power means that solar power can serve as a dependable core source of electricity as opposed to a niche solution, then it's pretty clear now as before that there is no reason to think we will have effective solar power in the foreseeable future.

'dependable core source of electricity as opposed to a niche solution'

5%-10% of total electrical power generated is not a niche solution. If you mean baseload when talking about 'dependable core source' then sure - but considering that no one (at least no one serious) has ever suggested that solar could take on a baseload role (see below for two obvious reasons why), that is a pretty easily dismissed concern. And by that definition, we will never have effective solar power ever, as you noted.

Lots of non serious people have created serious political support and expectations for solar.

Fair enough, of course. Yet the people in charge of Germany's electrical grid seem to be really quite serious, and doing a very good job. Unsurprisingly.

"Yet the people in charge of Germany’s electrical grid seem to be really quite serious, and doing a very good job."

The people in charge of Germany's electrical grid are charging their customers 3 to 4 times the US rate. They're doing the same kind of "very good job" that the US healthcare industry does. Slightly better for a whole lot more money.

It's become dated, yet little has really changed: "How Electricity Became a Luxury Good" in Der Spiegel. The English-language version is here:

'The people in charge of Germany’s electrical grid are charging their customers 3 to 4 times the US rate.'

And here is a complete breakdown of the amount paid by retail customers, as of Feb. 2017 -

'In February 2017, household power prices were on average comprised of:

Cost of power for supplier (19.3 %)

The profit margin and supplier’s cost of purchasing wholesale power on the market, some 5.63 cents/kilowatt hour (kWh).

Grid charges (25.7 %)

Charges for the use of the power grid, set by the federal grid regulator, some 7.01 cents/kWh.

Renewable energy surcharge (23.6 %)

The renewable energy surcharge pays the state-guaranteed price for renewable energy to producers and rose to 6.88 cents/kWh in 2017. [For further background on the renewable surcharge, read the CLEW Factsheet on Defining features of Germany’s Renewable Energy Act and the Green Energy Account]

Sales tax (value-added tax) (16 %)

The sales tax is 19 percent on the pretax price of electricity and makes up 16 percent of the after-tax price, some 4.6 cents/kWh.

Electricity tax (7 %)

A tax on the use of power, in Germany also called “ecological tax,” some 2.05 cents/kWh.

Concession levy (5.7 %)

A levy on the use of public space for power lines that the utility passes on to the consumer, some 1.66 cents/kWh, depending on the size of the locality.

Levy for offshore liabilities (0.1 %)

Grid operators must pay damages if they fail to connect offshore wind farms in a timely manner in order to sell the power they produce. Operators can pass these costs on to consumers through this levy. Due to higher chargebacks from previous years, the levy turned negative in 2017 and now stands at some -0.02 cents.

Surcharge for combined heat and power plants (1.5 %)

Operators of combined heat and power (CHP) plants receive a guaranteed price on the electricity they sell. The difference between the guarantee and the price they receive on the market is financed through this surcharge, some 0.43 cents/kWh.

Levy for industry rebate on grid fees (1.3%)

Large power consumers are partially or totally exempt from grid charges. These costs are distributed amongst consumers via this levy, amounting to around 0.388 cent/kWh.'

And from the same article, a bit of information that bears repeating - 'Despite being certain to see further price hikes, a stable majority of Germans support the Energiewende and consider it generally beneficial for the economy. A possible explanation would be that electricity consumed only 2.3 percent of households’ disposable income in 2015, up from 1.78 percent in 1998 and back to mid-1980s levels, before the liberalisation of the power market in 1998 lowered prices.' Again, to emphasize - over a 30 year time frame, the price of electricity in terms of disposable income is unchanged in Germany.

And hard as this might be believe - German gasoline prices are also 3 to 4 times higher than American prices. Which is undoubtredly why the German auto industry is doing so poorly, compared to the American one, right?

Not only is that dated, but this information 'In the near future, an average three-person household will spend about €90 a month for electricity. That's about twice as much as in 2000.' is simply incorrect, when contrasted with information from the recent past (see above about household disposable income) Just to emphasize that German retail energy prices simply haven't gone up that much depending on how one wishes to measure it - having currently risen to what West Germans were paying in the mod-1080s. Though wait until you hear about how much a movie theater ticket costs today in Germany, compared to 2000.

Was there any point to that long winded response? You spent a lot of time without directly addressing the actual point I made. Granted, that's the standard trolling tactic for you when confronted with any inconvenient fact. Here's the actual data:

Average rate of residential electricity in the US: $0.12/kwh

Average rate of residential electricity in Germany: $0.35/kwh

"And hard as this might be believe – German gasoline prices are also 3 to 4 times higher than American prices. Which is undoubtredly why the German auto industry is doing so poorly, compared to the American one, right?"

No, that's just about pricing the poor out of the automobile market to keep the highways free for the middle class and above.

‘Despite being certain to see further price hikes, a stable majority of Germans support the Energiewende and consider it generally beneficial for the economy. "

Germans are free to believe 2 plus 2 makes 5, but it's not going to change the math.

"And hard as this might be believe – German gasoline prices are also 3 to 4 times higher than American prices. Which is undoubtredly why the German auto industry is doing so poorly, compared to the American one, right?"

Sure, and they exported lots of them here, where gas is cheap. Again, many Germans may think overpaying is awesome, but the reality is it just makes them poorer.

>> And hard as this might be believe – German gasoline prices are also 3 to 4 times higher than American prices. Which is undoubtredly why the German auto industry is doing so poorly, compared to the American one, right?

Another post hoc ergo propter hoc fallacy from Clockwork. He's racking them up today.

To be fair to everyone, 5-10% of total capacity is not negligible and may be quite an improvement for high-insolation areas, but advocates need to be realistic about the long-term limits of solar, and how long it will take to get there.

Optimists are probably "triumphant" to use Tyler's word, because they have been besting pessimistic predictions for .. 30 years?

If I recall correctly, California as a whole consumes 8% solar and 9% wind on an annual basis. That's not negligible. It was also done so without a lot of the "gating" technologies that are often claimed to be necessary.

The fact is, there are a lot of big institutions (and small homes) which have peak power requirements during the day, and solar in the parking lot (or a nearby desert) can shave the peak without storage. That's where this growth has come from. I doubt the niche is exhausted.

Sure, to fully replace fossil fuels you need a lot of storage, but I think I can be optimistic about that on the appropriate timescale, say the next 100 years.

Glass half full, and gaining.

Yes, but....SW US is one of the better PV locations on the entire planet. Maybe amongst the best. Good all-year round insolation, and positively correlated with demand, as you say. We should expect it to do much better than the final global average.

Average global insolation for humanity is coastal china or India. Not so good.

I agree with a lot of what you and Battery Boy say. I view it is compatible with long term optimism.

The SW US has a lot of applications for solar .. price justified 5 years ago. The world will have more applications 5, 10, 20 years from now and until solar panel price per watt hits a wall.

"data from the U.S. Energy Information Administration showed California households paying 17.97 cents per kilowatt hour for electricity, or 40.9 percent more than the national average of 12.75 cents, according to the latest data from November 2016"

That 9% solar comes at a hell of a premium.

Like you though, I think solar has a good future. I've always been a fan.

Yeah, we have steep prices both to encourage conservation and to pay for older, poorer, renewables. And to pay for San Onofre cleanup.

San Diego Gas and Electric was at 35% in 2015 and is at 43% and growing now in terms of % solar/wind. Nearly the entire population of California has similar sunlight hours in a year. So I don’t see any reason why it couldn’t be near these levels for the state as a whole.’s-power-plants

In reference to the solar premium. San Diego County is at $0.12 per kWh for baseline residential users so not exactly a premium here.

The average (residential) electricity rate in San Diego is 37.63% greater than the national average rate of 11.88¢/kWh

1) I’m looking at my bill right now and my tiered rates aren’t what that site says.

This wouldn’t support your point even if we accept the number as true. The price per kWh is lower were the mix of solar is substantial higher. Suggesting the premium is driven by the non solar portion of production.

@Josh I live in San Diego and I just got my SDGE ( San Diego Gas and Electric bill for last month).
The 12.8 cents per KWh for baseline is for electricity delivery only; there is also an electricity generation rate added.

The total rate for baseline is 12.8 + 7.6 = 20.4 cents, the rate in tier 2 is 28.7 + 7.6 = 36.3 cents.

Counting various fees ,bond charges and local charges the effective rate per KWh is: ( its shown on the bill)

23 cents in tier 1, 40 cents in tier 2, 46 cents in tier 3.


Opening ANWR to drilling for oil and especially natural gas was an excellent move. Drill, baby, drill!

'Opening ANWR to drilling for oil and especially natural gas was an excellent move.'

To make fracking less profitable, that is.

Competition is wonderful thing. If a hellhole like Germany (still waiting for a new feckless grand coalition to form or a new election with the AfD managing to pull the rug out from under Merkel) wants to waste money lavishing subsidies on solar while burning more coal because of an irrational fear of nuclear power, have at it.

Hush. Don't mention the energiewende. Clockwork isn't concerned with maximising social welfare and the attendant hard maths. He just wants to virtue signal.

'Clockwork isn’t concerned with maximising social welfare and the attendant hard maths'

We have had a number of discussions about this - but please, do note who exports energy as the Energiewende occurs, and who imports electricty as they wait for their bright nuclear future to arrive - billions over budget and years behind schedule. Or are those maths too simple for someone to point out who likes to quote empirical data to buttress an argument?

The bizarre things is that the empirical data in terms of grid reliability and electricity exports shows that the Germans are much more interested in maximizing profit, after having done the attendant hard engineering. Here is the UK figure concerning exports, by the way - 'The UK electricity network is connected to systems in France, the Netherlands and Ireland through cables called interconnectors. The UK uses these to import or export electricity when it is most economical. In 2015, the UK was a net importer from France and the Netherlands with net imports of 13.8 TWh and 8.0 TWh respectively which accounted for 5.8 per cent of electricity supplied in 2015. Total net exports to Ireland amounted to 0.9 TWh.'


It's about total system cost-effectiveness, as you well know, and nothing to do with export or import totals or even production or consumption quantities (so no 'this year solar/wind produced xx% of demand' etc). If you want to make a social welfare argument (as do I) then let's stick to the right metrics, agreed?

It's my understanding that Germany both exports AND imports an unusually large amount of electricity as a result of the Energiewende.

'It’s about total system cost-effectiveness'

Or the will of the German voter to change the mix of electricity they are buying - somehow, that also gets overlooked in these discussions. Will it cost more? Of course. Did the Germans voters know that 15 years ago? Of course.

'so no ‘this year solar/wind produced xx% of demand’ etc'

Who cares? I have been talking about the amount of money that Germany makes from being a net electricity exporter and its grid reliability, which is empirical reality, after all. With data available for years that anyone can look at. Compared to the alarmists pointing out how the German electric imports will skyrocket and its grid will undoubtedly collapse by 2010, 2011, 2012, 2013, 2014 .... The google searching is fascinating to see all the strident warnings found in a certain section of the English language press. All of which have turned out to be laughably wrong as of January 2, 2018.

'If you want to make a social welfare argument (as do I) then let’s stick to the right metrics, agreed?'

The will of German voters seems just fine when talking about the German energy grid, actually. And they decided they would rather not use nuclear power, in exchange for paying more for their retail electricity, along with using more renewables. Though as it turns out, for German wholesale electric customers, the Energiewende has turned out to be quite agreeable. As noted here - 'German household power prices have reached a record high in early 2017 while wholesale prices are sinking. But despite the fact that Germans pay among the highest per-unit rates in Europe, their support for the Energiewende – the shift to a low-carbon economy – is strong. This is partly because electricity spending as a share of disposable income has remained steady for years.'

Of course, as pointed out in that article, as energy prices have increased in Germany, the Germans have responded in typical German fashion - 'Trademark German efficiency is also one way Germans cope with their energy bills. Like industry, German consumers have turned to savings strategies, like buying energy efficient appliances or switching to low-energy light bulbs. Over 20 years, German households reduced their power usage by 10 percent, while consumption in the United States increased by 20 percent. A German household in 2014 used less than a third of the power of an equivalent household in the US, and also less than other major industrial countries in Europe such as France, Britain and Spain.' Of course, the losers in this are the German utility companies. The Energiewende is also about efficiency, which when one is interested in selling ever more power, is a verboten concept. And no, this is not about virtue signalling on the part of German households, it is all about the hard cold cash.

"then let’s stick to the right metrics, agreed?"

That tactic is part of his trolling. He always moves the goal posts around or goes completely off topic rather than having a legitimate debate with pros and cons.

"Over 20 years, German households reduced their power usage by 10 percent, while consumption in the United States increased by 20 percent.

German electric prices are 3x what the US prices are. What would you expect? Forget this is an econ blog?

"A German household in 2014 used less than a third of the power of an equivalent household in the US"

Germans use half per capita, but your houses are almost half the size as well. Seems to be more of a standard of living issue you are having.

'That tactic is part of his trolling.'

Alistair is seemingly using cost alone, a metric repeatedly rejected by German voters who knew, before voting for it repeatedly, that the Energiewende would increase their electric bill. That is not moving the goalposts, that is reality. The people living in Germany simply have no interest in Alistsair's ideas of 'maximising social welfare.' They insist on using their own, oddly enough. And have actually punished political parties like the FDP for disagreeing with what voters wanted (admittedly, the FDP was a bit unlucky talking about expanding German nuclear power generation and jettisoning the Energiewende in the months before Fukushima occurred - causing the FDP to lose any role in the German federal government until the most recent federal election).

'Seems to be more of a standard of living issue you are having.'

Interesting way of looking at it, I suppose. Except most Germans who have travelled to the U.S. seem to actually feel their quality of life is better in Germany.


Yeah, one wants to engage in good faith, but it does look like Trolling. Clockwork keeps moving the goalposts and obfuscating the argument. Always dissimilate; never clarify. It's debate school 101 and he probably thinks it is terribly clever.

On the other hand, I have encountered Greenies who genuinely 'think' in the same disjointed way; incapable of systemic numerical reasoning but capable of an endless string of disjointed facts which they mistake for argument. So I'm not _absolutely_ sure Clockwork does it deliberately, or if he's just an educated fool without sufficient self-awareness and logic training.

'Always dissimilate; never clarify'

And here I am, providing citations and links to information. This whole thing started with the observation that for cell phone and lantern batteries in 'emerging economies,' a solar panel is more than sufficient. And somehow, the idea that for a rural villager in India or Africa or South America a solar 40W or 60W or 80W panel has no imaginable drawbacks related to power storage, we quickly change into another discussion entirely. Talk about changing goalposts - here is your response to the idea that a solar panel would easily cover a villager's need to recharge cell phone and lantern batteries - 'The problem is the duration of storage. Everyone expects you just need to store the energy for a day or so (right?). But in reality, for a near-100% renewables grid, you need to store it for months, not hours.'

The entire point of my first comment is that the rural villager has no need for a grid, in the same way that a rural villager using a cell phone has no need for a copper wire where they live that runs to a switching station to make a phone call

We go from 'emerging economies' not actually needing a grid for rural villagers to positing something that will never happen anyways - a near-100% renewable grid in an industrial nation. (Leaving aside powersats as a concept.)


>> And here I am, providing citations and links to information.

Ah yes - you repeatedly post long excerpts and links to data which are irrelevant to the argument. For which you are repeatedly called on. You do this 2 or 3 times in a row then when pressed claim the argument was about something else altogether. You never reference formal syllogisms, and commit a variety logical fallacies, though you obviously have some training in debating skills.

The only reason I don't entirely believe you are a Troll is that I've met too many Greenies who "think" in this way. All facts and no capacity for logical argument.

"So I’m not _absolutely_ sure Clockwork does it deliberately, or if he’s just an educated fool without sufficient self-awareness and logic training."

I agree that it's difficult to tell if he's an intentional troll or just incapable of any kind of rational debate. However, from a practical point of view it hardly matters. If he was interested in changing his behavior he'd at least make the attempt to argue in good faith. He doesn't do so and since I can't determine his intent, I'll respond to his observed behavior.

It's definitely not trolling.

First, 'socialist hellhole' is the catchphrase.

Second, the solar subsidies were capped several years ago, by that selfsame feckless Merkel government.

Third, the rational fear of all the costs associated with getting out of nuclear is driving the nuclear industry's increasingly desperate attempts to keep those costs out of view in Germany - mainly because it is the German taxpayer on the hook for those costs, not the industry that profited from causing them. And of course, Germany is missing out on the nuclear power benefits that the UK will be enjoying any decade now - 'Two years before construction work begins in earnest on the UK’s first new nuclear power station in three decades, costs are already beginning to rise

French state-owned electricity generator EDF has warned that the cost of building Hinkley Point C nuclear power station will be up to £2.2bn higher than its last estimate, reaching as high as £20.3bn. Construction work on the plant’s ancilliary facilities has already begun, but the first concrete for the building that will house the reactor itself is scheduled to be poured in 2019. EDF says that there is a risk that the site’s first reactor, due to come on-line in 2025, might be up to fifteen months late, with the second unit possibly running nine months behind schedule.

Of the increased cost, £1.5bn results from “a better understanding of the design adapted to the requirements of the British regulators, the volume and sequencing of work on-site and the gradual implementation of work on-site.” The scheduler overrun gives rise to the remaining £0.7billion, the company adds.

This represents the second increase in the cost of the project, after EDF raised the estimate from £15bn to £18bn in 2015 because of inflation.'

The Finns are so enthused with nuclear power that they will apparently be waiting another 5 months to enjoy its benefits, while importing power from countries that are net electric power exporters - 'Finland will have to add another five months to the decade-long wait to start production at a nuclear reactor once billed as the world’s biggest.

Olkiluoto-3, able to power about 3 million homes, will be delayed until May 2019 from its previously expected to start at the end of next year, according to Teollisuuden Voima Oyj, the Helsinki-based utility that will operate the unit. Areva SA, the supplier, said it needed more time to adjust the production schedule.

Plagued by cost overruns and legal tangles, the delay is the latest setback for the 1,600-megawatt reactor meant to provide cheap power to companies from Finnish papermakers UPM-Kymmene Oyj and Stora Enso Oyj to a Google Inc. server farm. The nation last year had to import a quarter of its electricity in the wholesale market, where prices for delivery to Finland are about 21 percent higher than the Nordic-region average.'

And how did the socialist hellhole that is Germany do in comparison to Finland's need to import electricity while waiting for its bright nuclear future?

'Electricity Exports by Country

Below are the 15 countries that exported the highest dollar value worth of electricity during 2016:

1. Germany US$3 billion (11.5% of total electricity exports)
2. Canada: $2.2 billion (8.3%)
3. France: $2.1 billion (8.1%) -

(That is net export, one should note, and not the sort of cheap shell game that makes a net oil importer like the U.S. into an oil 'exporter.')


Are you parents happy with the way your life has turned out, or are they disappointed?

Not to get too detailed about it, but yes, both parents are proud with how my life has turned out. But then, I also got a varsity letter as a sophomore, am an Eagle Scout, and done a number of other things not worth mentioning here.

They even came around to my riding a motorcycle for the last three decades.

Yes, Germany is a socialist hellhole and the alternative (AfD) would continue making Germany a national socialist hellhole.

If Germany were serious about meeting its carbon emission targets, it would gladly shift some of the subsidies that were going to solar to subsidize nuclear power instead of burning coal and gas.

That's too bad; something seems to have gone very wrong for you somewhere.

I agree with opening ANWR. And I think that the oil found there can be extracted at a reasonable price. But I don't see much of a market for the natural gas. The gas would have to be converted to LNG to be shipped and that is expensive. The current market price for natural gas is about $3 per 1,000 cubic feet. Liquefying the natural gas and shipping it on an LNG vessel ship to Asia would be $5-6 per 1,000 cubic feet of additional cost. I think solar would be brought on board pretty quickly at that point,

Natural gas works well with wind or sun because you can turn a natural gas plant on or off like a lighter. A coal plant uses steam technology and bringing one up requires waiting for the water to boil to create steam.

'have already seen solar energy flatten out in the range of 5 percent to 10 percent of total energy use'

As compared to 20 years ago, when solar represented essentially 0% of total energy use? And basically no one has ever believed that solar energy could ever be more than something like a third of total, at the outer practical extreme. The reason why is left up to the reader - or is that solve for the meteorology and orbital mechanics equilibriums? And of course, there is also an absolute limit to the amount of power solar energy can deliver - starting with a theoretical realistic maximum of around 1050 W/m2, and going down from there -

'Also note that about 95 percent of global energy storage capacity is from hydroelectric power'

With a not miniscule percentage of that likely involving pumped storage, which is completely agnostic as to where the electricity to run the pumps comes from - nuclear (as originally designed), wind, solar, etc.

And from the column - 'There is now a doctrine of what I call “solar triumphalism”: the price of panels has been falling exponentially, the technology makes good practical sense, and only a few further nudges are needed for solar to become a major energy source.' It has been clear for a while that the Chinese have been price dumping on the global market for solar panels - those low prices are a sign of a distorted market, and play a role in wiping out R&D and suppressing 'support and entrepreneurial and policy dynamism.' Not that one would expect this web site to highlight such things.

And living in Germany, this is going to require a bit more than bare assertion to be credible - 'Germany and the state of California have experienced operational problems as solar has grown as an energy source.' If one means technical challenges, sure. If one means that the German electric grid has declined to the level of the U.S., then absolutely not - as can be seen here, where the German grid reliability is broadly improving as the percentage of renewal electricity in the mix has risen

Unsurprisingly, comparable American statistics are difficult to find presented in such a clear fashion, but this EIA info is at least indicative of how well the U.S. does in comparison - 'In 2015, municipal utility customers experienced the lowest instances of power outages in both frequency and duration, averaging one outage and about two hours of interrupted service. With major events factored in, investor-owned utilities' customers averaged slightly more than three hours without electric service, while co-op customers averaged nearly five hours without power. Co-op customers, on average, experienced about twice as many outages as customers of investor-owned and municipal utilities.'

I suspect that the route to more solar power than today lies more with manufacturing and installation innovations not really basic physics. The internal combustion engine is an incredibly complicated and low efficiency technology, but what has kept it in the running is literally millions of tiny innovations and tweaks in design and manufacturing plus huge scale. It is very difficult to know if there is the same opportunity for solar but I wouldn’t bet against it. A physicist and solar «expert» is probably not the right guy to provide an opinion on this. We have had some recent threads on how we don’t really understand innovation, and the 1% R&D doesn’t frighten me, the kind of innovations needed now never show up in the R&D budget.

My view is that solar still has some way to go, especially in equatorial countries like India, but that the final solution will eventually be nuclear.

I cannot help but be both brightened and frightened by your last sentence!

It's the only way to be sure.

+1 for Aliens

Game over man, game over.

Although solar and wind power may not provide electricity all the time, they can provide it some of the time, and this is reducing the emissions from fossil fuel burning.

Battery technology can smooth out short term fluctuations, but does not resolve the problem of long periods of cloudy wind-less weather.

Yes, BUT: Firstly, Solar PV and wind are not carbon-neutral to build. You have to account for lifetime carbon costs. Secondly, even with good wind models, you have to have need some gas as spinning reserve and short term reserve, which runs very inefficiently. Thirdly, you incur more carbon costs through any storage costs (offset against reserve, above). Fourthly, you incur more carbon costs through higher grid and connection costs for the renewables.

All of these factors reduces the carbon saving somewhat. You save about 2/3rds of the fossil alternative as a reasonable rule of thumb, maybe.

'Firstly, Solar PV and wind are not carbon-neutral to build.'

Nothing is carbon neutral to build.

'Fourthly, you incur more carbon costs through higher grid and connection costs for the renewables. '

Really? You mean solar panels on a factory roof (which also reduces line loss when providing power) cause a carbon cost through their grid and connection costs - please do explain, and don't be afraid to use the attendant hard maths.

>> Really? You mean solar panels on a factory roof (which also reduces line loss when providing power) cause a carbon cost through their grid and connection costs – please do explain, and don’t be afraid to use the attendant hard maths.

Me, (and lots of other analysts) already have. Solar Panels on roofs are a stupid idea (at least on a large scale). Basically, the additional installation and maintenance costs (plus worse-aspect) outweigh the savings on land and grid costs, by a factor of about 2. Building on roofs is about 50% more expensive overall, per unit of installed capacity. This factor rises as solar PV gets cheaper.

If you must do large-scale solar PV, do it on a open field. There, I've explained, and kept the maths simple for you.

'Basically, the additional installation and maintenance costs (plus worse-aspect) outweigh the savings on land and grid costs, by a factor of about 2.'

A whole bunch of extremely cost conscious people in this region are going to be shocked to hear that their experience is contradicted by your words. Including Aldi, which is a company noted for its extreme profligacy.

'Building on roofs is about 50% more expensive overall, per unit of installed capacity.'

As noted by the Aldi example, it seems as putting panels on a roof as part of the process of building the roof is not all that expensive - though as also noted, Aldi is well known as a profligate company.

'There, I’ve explained, and kept the maths simple for you.'

Except for the part where you explain how one incurs more carbon costs through higher grid and connection costs, as the carbon 'cost' is a distinct entity from grid and connection costs.

Well, that's the cost data if you want to refute it. Do we need to go to sources?

Note that you can't use anecdotal examples like that as a refutation for a social welfare argument. You have to show costs. It can be perfectly cost-effective for individuals to mount roof panels for a variety or reasons (like subsidy farming) but it still remains LESS cost-effective than putting them directly in fields.

Look, clockwork, sometimes you come across as a well-educated type, and sometimes as a complete troll. Extending the principle of charity, I'd suggest that a lot of our disagreements comes because your arguments are badly composed, not that your basic facts are wrong. As one analyst to another, you need to have a better quantitative system level model of costs, and a clearer distinction between effective and cost-effective in argument.

Aldi is a generally considered a close mouthed organization, but let's see what comes up.

First result, German only -

In Germany, Aldi Süd is the largest private operator of rooftop solar installations. In 2016, they produced 114 million kilowatt hours. Aldi covers about 15% of its electrical use through rooftop PV, tendency upwards. Whenever a new store is planned, the PV installation is part of the planning (assuming conditions are right - a ground floor Aldi inside a multistory shopping complex is unsuited). Aldi currently uses 80% of its self-generated power internally, and would like to make that 100% - and is looking for future oriented storage solutions because the 20% they do not use is generated during periods of peak solar irradation. Three stores in Frankfurt are currently testing a 100kw battery system, in combination with EnBW. Aldi has invested in energy saving technology also, including certified energy management systems. All Aldi branches have energy monitoring systems to record energy consumption.

Hopefully, that is a bit of not anecdotal data from what is essentially half of Germany's largest discount chain (Aldi Nord is the other half), both of which are famous in Germany for 'turning pennies twice' before spending them. Aldi is equally famous for never doing anything that does not make it a profit, though as it is a private company, the time horizon for that ROI is both secret, and determined by Aldi itself, and no one else.

"Aldi is equally famous for never doing anything that does not make it a profit, "

Do they advertise?


Anecdotal counter-example isn't refutation for statistical argument in general and is also a fallacy of Denying the Antecedent in this particular.

Doubling down on Aldi doesn't help you - it just makes you seem incapable of understanding logical structure.

' Do we need to go to sources? '

Of course - I put links to just about all the data being cited. Though it is true that actually providing sources or links seems to be the sort of thing that commenters here don't consider important when making comments, and have no problems ignoring when provided. But yes, please do post the source(s) for 'by a factor of about 2. Building on roofs is about 50% more expensive overall.' Because from what I've seen of rooftop installations on new construction in Germany, it is hard to imagine that it costs even 5% more - the rooftop frames use considerably less metal than the solar farms (like the ones near SAP), not to mention not requiring any ground anchoring, and the expense and manpower of screwing them into those frames is not different. That anchoring is quite relevant, by the way - panels flat on a roof do not require anywhere near the digging/footings of standalone frames. Which have to withstand over 100kmh winds, like what much of this region experienced today

'like that as a refutation for a social welfare argument'

You really seem to think money=social welfare. German voters don't agree with that framing, and haven't for at least a generation.

'Anecdotal counter-example isn’t refutation for statistical argument in general '

Again, please show your sources.

There you go. I'd call your attention especially to the Bloomberg projections vs similar costs for industrial greenfield sites. The engineering costs dominate in the long run as module costs fall.

One thing I do enjoy about your posts is how they reveal the limits of your systems thinking and focus on the wrong details. Metal Frame costs are negligible compared to inverters and wiring etc. And flat roof panels lose far more in efficiency than the saving in frame costs warrant. They also silt up faster and require more cleaning. And I note you're only talking about ideal flat and structurally robust roofs, presumably with elevator access. Of course, most rooftops don't have that and the real costs reflect that.

Anyway, that's fun enough for today. Brown-out, as I always say....

>> Nothing is carbon neutral to build.

Of course. And it varies by base energy mix. The proper way to do the analysis is not to target carbon saving at all, but simply put a price on carbon and minimise total cost for given constraints at the system level. This will allow direct comparison of the cost of different options and incidentally tell you the net carbon saving.

The point is that renewables carry much more of their lifetime carbon in initial capital phases rather than operating phases, and those need to be properly discounted when calculating "savings".

Lithium Ion may not be the ideal battery technology, but it has nothing to do with the inability to store energy long-term. There will never (at least in out lifetimes) be a business model that supports long-term energy storage at any scale.

Solar and battery storage business models depend on near 100% utilization rates as there are no marginal costs. Storing energy for two days at a time instead of one, effectively reduces the utilization rate of the battery in half.

Lithium Ion is ill-suited to grid-based energy storage for a variety of reasons, discharge limitations (depth of discharge and discharge related deteriorated) being among the most important. Over time there are likely to be a huge amount of incremental improvements to battery production from applications of more appropriate technologies. But long-term storage capability is basically irrelevant in most applications.

Prospects for improvements in solar is actually a completely different situation. Solar panels already appear to be able to produce less expensive electricity in a wide range of applications (basically anywhere with good year round insolation). Technology improvements may make economics better, but won't address the fundamental problem that they can only produce electricity when the sun is shining - and it will never shine at night or when it is raining (which happens for months in the tropics)

'There will never (at least in out lifetimes) be a business model that supports long-term energy storage at any scale.'

Ever heard of pumped storage? Like this place, which is now 40 years old? - 'The Bath County Pumped Storage Station is a pumped storage hydroelectric power plant, which is described as the "largest battery in the world", with a generation capacity of 3,003 MW The station is located in the northern corner of Bath County, Virginia, on the southeast side of the Eastern Continental Divide, which forms this section of the border between Virginia and West Virginia. The station consists of two reservoirs separated by about 1,260 feet (380 m) in elevation. It is the largest pumped-storage power station in the world.'

And as pointed out above when talking about pumped storage, that facility does not care in the least where the electricity that drives its pumps comes from, and the infrastructure already exists. Seems as if that generation old business model is quite functional - maybe you were unaware of it, as compared to saying it will never happen in the future.

Sorry Clockwork, you are just plain misunderstanding the problem.

BB is pointing out that utilisation rate matters for cost-effectiveness of storage. It's nothing to do with storage cost per se or the storage technology. It's about how low utilisation rates depresses the return on investment. For large-scale grid storage (>30% renewables), utilisation rates are LOW (singe figures per annum to smooth seasonal variation). It doesn't matter if it is pumped hydro, batteries, or anything.

Suppose you have a storage tech which is good for 5,000 storage cycles. It matters a great deal whether those 5,000 cycles are expended over the course of an afternoon or over 25 years. In the first case you get your returns instantly. In the second case you have to wait 25 years for your returns. Which is the better investment? The effect of low utilisation on return on investment is very large.

Storage is a way of transporting energy through TIME. Your storage technology is like a ship carrying cargo; the lower the utilisation rate, the further those ships have to travel to deliver their (average) cargo and the less cost-effective they become.

Correct. Pumped storage charges and discharges daily.

The facility may last for a long time, but it stores energy short-term.

So, there is a business model? Who would have guessed. Well, maybe anybody familiar with pumped storage, which is a concept stretching back to the introduction of large scale nuclear power generation, to deal with the fact that a nuclear plant is essentially unable to throttle its output to respond to changes in demand throughout the day. The sort of technical problem that undoubtedly makes nuclear power completely unsuited in an electrical grid, right?

It rather proves BB's point, Clockwork, utilisation rate matters.

Pumped hydro and nukes couple on a diurnal cycle. They charge and discharge once a day. It wouldn't be economic to build the pumped hydro if they only charged and discharged once a month.

'It rather proves BB’s point, Clockwork, utilisation rate matters.'

Pumped storage facilities in Germany will not be going away when the last nuclear plant goes offline in Germany. Nobody is then going to abandon such facilities, in part because they already exist, and in part because it allows excess power generated by renewable sources to be stored until needed.

The tight coupling with nuclear is true. And not necessarily relevant when the storage facility is used to compensate for other patterns of power generation.

'It wouldn’t be economic to build the pumped hydro if they only charged and discharged once a month.'

Sure, but the idea that a pumped storage facility needs to be fully charged and discharged is part of the nuclear coupling. There is no good reason reason preventing such a facility being partially charged and discharged all through a 24 hour period, to balance out shifts in the weather, for example.

One can assume that the people involved in this 2017 project are not precisely thinking about a diurnal cycle (though that comes with the PV territory, of course) - 'A coal-mine that powered German industry for almost half a century will get a new lease on life when it’s turned into a giant battery that stores excess solar and wind energy.

The state of North-Rhine Westphalia is set to turn its Prosper-Haniel hard coal mine into a 200 megawatt pumped-storage hydroelectric reservoir, which acts like a battery and will have enough capacity to power more than 400,000 homes, said state governor Hannelore Kraft. The town of Bottrop, where people worked the 600 meter (1,969 foot) deep mine since 1974, will keep playing a role in providing uninterrupted power for the country, she said.

Germany’s decision to turn a coal mine into a pumped-hydro-storage station may solve two of the most intractable challenges created by its shift to clean power. On a local level, it provides new economic activity in a region where generations of workers have relied on fossil fuel for their livelihoods. On a regional level, it catalyzes the expansion of renewable energy by helping to maintain electric capacity even when the wind doesn’t blow or the sun doesn’t shine.'

And Germany is not the only place thinking of using an old mine for pumped storage - 'Dominion is looking at building a pumped storage facility in Southwest Virginia. It has focused its search on two sites — the old Bullitt mine near Appalachia and a Tazewell County location. Either project would be smaller in scale than the Bath County site, which is the world’s largest pumped storage facility.


The Bullitt site is in the conceptual phase, with researchers from Virginia Tech evaluating it, said Dominion officials. The upper reservoir and powerhouse would be similar to the Bath County site. The biggest difference would be that instead of a lake-style lower reservoir, the water would be stored in the old mine works.'

Oddly, it seems when people start looking for solutions for an 'intractable' problem, such as only a limited number of sites suited for pumped storage, answers can actually be found that expand the definition of 'suited.' Because there are likely a number of mining sites that could fit this description - 'When needed to compensate intermittent wind and solar power, as much as 1 million cubic meters of water could be allowed to plunge as deep as 1,200 meters, turning turbines at the foot of the colliery’s mine shafts. The mining complex comprises 26 kilometers (16 miles) of horizontal shafts.' Probably even in the UK, with its long history of coal mining.

And once again Clockwork generates a lengthy post with excerpts which totally fails to engage with the actual claim by BB about utilisation.

Or you are misunderstanding that he wrote 'a business model that supports long-term energy storage at any scale.' Which is empirically simple to prove as being wrong. Particularly in light of the very second paragraph from the wikipedia article - 'Construction on the power station, with an original capacity of 2,100 megawatts (2,800,000 hp), began in March 1977 and was completed in December 1985 at a cost of $1.6 billion. Voith-Siemens upgraded the six turbines between 2004 and 2009, increasing power generation to 500.5 MW and pumping power to 480 megawatts (640,000 hp) for each turbine. Bath County Station is jointly owned by Dominion Generation (60%) and FirstEnergy (40%), and managed by Dominion.It stores energy for PJM Interconnection, a regional transmission organization in 13 states and the District of Columbia.' Somebody seems to have thought that paying for that upgrading made business sense - a decade ago.

But each storage cycle is short! Utilisation is high and keeps the RoI reasonable. They don't store the water for months. Don't you understand what is being said here?

'They don’t store the water for months.'

Of course they don't currently. See above for ideas of how pumped storage might be used in the future, as compared to today. Including the idea of what in German is called an 'iron reserve' - that amount of safety margin you only use when things are outside of normal boundaries.

Which is already the case in the current electrical grid - some power plants (mainly obsolete ones) have an exceeding low current utilization rate - they are there to provide a back-up, whether for emergencies or to handle scheduled maintenance at other plants. Oddly enough, there are a fleet of obsolete coal plants with a close to zero utilization rate being maintained in Germany right now, just in case they are required for unanticipated reasons, encompassing 8 plants that can generate 2,700 megawatts. Which essentially everyone involved hopes will not be used, actually. But which the customers get to pay for anyways, to the tune of 780 million euros a year.

Alistair is correct. As per my comment above "The facility may last for a long time, but it stores energy short-term".

I thought it was fairly obvious when I said that batteries would never store energy at scale for more than one day at a time that I did not mean that we would be replacing batteries daily.

If you want to make your point, you would need to show me that the average time that any pumped storage facility in the world stores energy is longer than 24 hours.

I think you will find that the business model for pumped storage in (or compressed air, batteries, fly wheels, etc.) requires daily charging (storage of water) at times that electricity is in over-supply (night in the case of nuclear, midday for solar) and discharge (releases of water) when it is under-supplied.

At the scale of pumped storage this requires differential pricing for electricity at different times. Batteries in certain applications can have other revenue streams such as reduction of demand changes and other grid services.

'I thought it was fairly obvious when I said that batteries would never store energy at scale '

There are other ways to store energy in a 'battery,' which is one of the terms used to describe the Bath County pumped storage facility.

'If you want to make your point, you would need to show me that the average time that any pumped storage facility in the world stores energy is longer than 24 hours.'

Well, then we will need to wait until the last nuclear power plant is turned off in Germany, and then see how pumped storage is used. As a guess, the way you are looking at it will be irrelevant - when there are days of excessive power generation, all pumped storage facilities will be recharged. And as long as that excessive power situation continues (winter storms lasting several days, extremely sunny days in a row, to note two examples that have occurred in the last couple of years in Germany), the pumped storage facilities will be storing energy until needed. And one assumes that though full charging may occur, the main use of pumped storage will then be averaging out daily fluctuations - neither fully discharging or recharging. Or perhaps keeping a reserve of something like 10-25% during the winter, to help handle possible cold spikes - would that count as long term storage if such a scenario comes to pass?

One can laugh all one wishes at the term Energiewende, but a major component is changing the framework of how an electrical grid is looked at, instead of using ideas that rely on top down centralized planned solutions from decades in the past.

And +1 for that. ROE on batteries is dominated by utilisation. But those would be very low for large scale deployments.

There is actually quite a large business that supports very large scale and long term energy storage (with an annual storage-release cycle) - underground natural gas storage. This occurs worldwide. In the US ...

"The principal owners/operators of underground storage facilities are interstate pipeline companies, intrastate pipeline companies, local distribution companies (LDCs), and independent storage service providers. About 120 entities currently operate the nearly 400 active underground storage facilities in the Lower 48 states.

Owners/operators of storage facilities are not necessarily the owners of the natural gas held in storage. In fact, most working gas held in storage facilities is held under lease with shippers, LDCs, or end users who own the gas."

Good point. What's the utilisation rate though?

Who cares - the main reason that such storage exists is because the demand for natural gas in the winter cannot be met by direct feeding from pipelines. The amount of gas flowing from wellheads isn't sufficient either. Storage is used to ensure enough natural gas is available during its peak use. Seems as if someone solved one of the most insurmountable problems of using natural gas - that is, it just cannot be produced and transferred from production to consuming regions via pipeline in adequate quantities to handle such seasonal peaking. Of course, as we all know, such technical problems are utterly insurmountable.

(There is a large facility just down the road from GMU, though most people are completely unaware of it - the Ravensworth Station reservoir.)


>Who cares

Remind me never to let you near asset management. You have no sense for inert capital.

In his column, Cowen mentions the opportunity costs of solar (i.e., the very large investment by power companies to produce electric power by using conventional sources such as coal, oil and gas, and nuclear). Cowen doesn't mention power companies' political influence with state legislatures, which have adopted laws that make it difficult or impossible for small producers of solar power (e.g., individual homeowners) to transfer excess solar power back to the power company as an offset to the charges of the power companies. Although Cowen typically supports the "disruption" that comes with technological innovation (e.g., solar power) and opposes efforts by owners of old technology (such as power companies) to discourage innovation, for some reason that isn't reflected in his latest column at Bloomberg. Cowen doesn't use the term in his column, but one might get the impression that the "sunk costs" in power plants using old technology are so great that a major shift to solar would cause more than a little "disruption". My observation is that solar can be a significant supplement as a source of power but not a replacement for conventional sources, and that for solar to be a significant supplement requires the cooperation of the power companies, which is about as likely as Donald Trump reaching out to Sen. Warren for cooperation in crafting legislation to fix Obamacare.

Sure, but at the same time, there are negative externalities here.

Those small suppliers are free-riding on the reliability of the grid to supply them when their local solar is down. This cost is borne by other regular consumers who effectively subsidise them. Also small scale PV generators tend to receive a fixed (average) price for their output when the actual spot price of power during period of peak solar PV is very low.

Ii think the home owner tariff is a bit of a red herring. These are tiny, tiny energy producers - the most expensive and least efficient ones on the table.

I note how rare it is for larger commercial buildings (warehouses, malls, big box stores, etc.) to have solar, either new-build or retrofit. If its not the default for a 100,000 sq ft new build warehouse to put solar panels on the roof, I'd say the economics aren't yet compelling.

There are a few exceptions - Walmart for instance has put panels on a lot of their stores, but I suspect that is driven more by public relations that economics.

Not all big buildings are big enough consumers to justify production for own demand. Hospitals and high schools rather than warehouses.

My point is that if rooftop scale solar is really an economically attractive idea (as opposed to very large utility scale), we should see it on sites with largest rooftops. One might say household, commercial, and utility scale. Even if the 250,000 sq ft warehouse consumes little or none of the power produced on its roof, it would seem someone should find it attractive to put the panels up - the utility, a 3rd party, the warehouse owner.

If that's not standard practice, a few excess watts from a house seems the wrong thing to be concerned about.

I would agree, however, that there is additional robustness (i.e. civil defense) value is having hospitals and high schools able to generate at least some of their own power. IIRC, there are about 20,000 high schools in the US, already carefully matched to population. At least where I live, many of the school roofs have unobstructed sky views.

For robustness to be improved, the solar power has to be able to disconnect from the grid in the even the grid is down. This is possible, but not common with household systems. The default is to disable the solar power if the grid is down.

I believe that the deal our local power company offers is not favorable to trading, so you have to be really sure you can consume your power when you build your array.

That is the difference between a hospital and a warehouse.

A hospital can look at its electric bill, the cost of a car park solar array, and run the math.

A builder of commercial property, with uncertain future energy requirements, can't do that. Missing data.

For five years before I retired in May 2015, I worked in the Empire State Building (ESB), NYC.

How many acres of solar panels would be needed to electrify the ESB? Aside from Central Park and other parks, there is very little acreage available in Manhattan. If there were, it would sell for tens of millions of dollars an acre. On a national level, would there be any remaining acreage to grow food?

Anyhow, it's been politicize and effective green graft opportunities aren't diminishing.

This man is not an engineer.

What you do first is an energy audit and efficiency drive. I expect this has been done, and the building is running at about half the energy per square foot it was at peak.

Then you look at the city's energy supply mix as a whole, and see what you can offset with local been sources, like say

You make the transitions over time as both efficiency and generation technologies come on line.

Meh, it would be trivial to build solar farms in existing pasture land in upstate New York (paying a higher lease rate than the rancher could get for meat / milk production). Indeed, most of the country has excess farm land, since the productivity per acreage has increased drastically over the last century.

Available land isn't the bottle neck for solar/wind. The bottle neck is storage costs.

+1. Land costs are slight.

I'm not sure they will ever be compelling, at least at a societal level, without subsidy.

As above arguing with Clockwork, the savings per installed KWh in land costs are overwhelmed by additional installation and maintenance costs. Land costs turn out to be surprisingly quite a small component of overall costs for solar PV.

This is a technology problem. As the cost per W of solar cells (or whatever the future technology will be called) continues upon its exponential decrease, more and more research will be focused upon installation technology (eg: cheap roofing material that generates power, cheap bricks that generate power) at which it will likely be economically viable. Heck, maybe they even put a nano super cap into the roofing for storage.

My point is not that solar is competitive today, it is not, but that the cost curve has a great history and is quite powerful.

Umm... Rhubarb. Sorry. There are no strong reasons to assume solar installation costs will fall faster than solar hardware costs (especially when regulatory effects are included!). Indeed; the clear historical trend is in the OTHER DIRECTION; installation and maintenance have been growing components of total system cost.

The technologies you mention seem to be more far apt at reducing hardware costs than installation. If I have a cheap PV cell, why fuse it into an expensive brick and integrate it into a building when I can mount it on a plastic scaffold in a field? At least I can replace the plastic scaffold easily when it wears out too. Honestly, try to think a little bit beyond the "gee whiz" level and think about the system costs.

The argument was not solar vs no solar. It was solar in fields vs solar on rooftops. Solar on rooftops will probably always be less cost-effective than in solar in fields.

'installation and maintenance have been growing components of total system cost'


“Indeed; the clear historical trend is in the OTHER DIRECTION;” — this was the point.

The exponential decrease in the price of PV has almost made it competitive. I expect that entrepreneurs will now focus upon the installation side of the problem.

I often read that the price for solar panels has dropped significantly, in particular for the panels made in China. Of course, lower price often means lower quality. And where are solar panels likely to be used? Not in North Dakota. Solar panels are most likely to be used as a source of power in the south and southwest, which have hot climates that reduce the life of solar panels, especially those of lesser quality. Creating tax incentives (credits) to install solar panels made in China is, well, not in a politician's best interest. Again, solar can be a great source as a supplement to power generated conventionally, but not as a reliable alternative source of power, at least not in America where power company customers go ballistic if power is cut off for a few hours during a blizzard or hurricane. Americans expect power to be available, and available cheaply, all the time, which is an expensive proposition. Those who reside in hurricane prone areas know what I mean. If you don't, then you cannot imagine the army of power trucks and personnel, from far away (including North Dakota), that descend on areas hit by hurricanes within hours of the catastrophe, working round the clock to restore power. Rather than getting worked up over solar, I would instead devote the resources to getting power lines under ground. Old photos of towns and cities soon after electricity was available, with power lines criss-crossing streets, aren't that much different from today. Rather than devoting resources for a more efficient method for getting power to the user, the power companies cut down trees that might fall across power lines, which reduces shade and increases power usage. The only innovation I have observed in my community is the installation of digital meters, which has accelerated billing of customers. As between underground power lines and faster and more efficient billing, power companies choose the latter. Econ 101.

I live in a neighborhood built 20 years ago, with underground power to the houses. Of course, power to the neighborhood is above ground. Underground power lines are a popular topic after every hurricane. In the end, it comes down to cost - underground costs considerably more to install and maintain.

Digital meters do improve billing (cost savings, more accurate), but also help considerably with system diagnostics (detecting outage locations). Some services can be performed remotely and don't require a truck roll (again saving money, and reducing response time.)

And at least with my meter, I can see power use in 15 minute increments, which has been useful in doing power usage analysis. Longer term, I think a big driver to to support dynamic pricing - my utility now offers rate plans with cheaper power at night and on weekends. I suspect finer grained power pricing options are probably in the future. All in all, I think its a good thing.

Sure, but as Alistair correctly points out, the small solar providers actual contribution to reducing grid costs is nearly zero, not whatever the state charges them per kwH.

Best case, if the state allows trading, what you end up with a system where power is really cheap, even zero-cost, at peak solar production (i.e. sunshiney days around noon), but unchanged at other times. But the utilities' ability to selectively switch off demand is pretty limited, esp. in terms of how much they can save that way given fixed costs.

The lack of R&D amongst solar manufacturers is interesting, but may have limited relevance.

If prices determined by recent auctions are meaningful, then solar in many applications is so much cheaper than any other option that there is no barrier to expansion that would be addressed by improved performance.

Higher performing (or less expensive) panels do make it possible to spend money on other system elements that facilitate expansion (batteries, transmission, etc), but don't address the fundamental problem that when the sun isn't shining the panels aren't producing.

I say if the auction prices are meaningful because I can't figure out how developers and investors aren't losing money. There are no reasonable inputs that I can stick into a financial model that makes the economics of solar work at $0.21/kWh as seen in Mexico.

The article suggests that the low prices will be dealt with by cutting costs in installation (at the risk of quality). However, I suspect that climate fever is causing firms and investors to abandon common sense and do anything to get into projects. I can't see how anyone is going to make any money on these projects and in fact don't really see how they will pay back debt. But would be love to hear back from anyone who knows something that I don't.

I know someone in an oil company charged with "buying green technology" and they are throwing money at it pretty hard.

The key question is "how much faster could solar actually grow?"

Infrastructure has both long construction cycles and long life. Our local gas fired plant was built in 1958, ran until 2000, when a project was begun to replace it by 2020. I guess it's on track.

You have to be a bit outside the engineering community to think things like a transition to solar can't happen "now" because whatever. Nothing happens in the now. It usually happens on a 20 year cycle, or longer.

In this 20 years we are building a lot of solar and wind because we can consume the output immediately. No futher invention required.

Excellent Blog! I would like to thank for the efforts you have made in writing this post. I am hoping the same best work from you in the future as well. I wanted to thank you for this websites! Thanks for sharing. Great websites!

Oh, wow. Can I get this for California?

I fear our regulators would be conflicted on safety vs environment.

In short, the optimists are optimistic. Fine... now let's get on with working through the problems.

And here is a crazy idea .. renew NREL funding.

Spain's flattening out of solar has a very clear explanation: When the Rajoy government came in, they removed all subsidies for solar. It won't surprise anyone that electric companies strapped for cash at the time would stop all their plans for new installations: Spanish electrical companies are not exactly known for their love for risk and investment.

"Is effective solar power further away than we had thought?" Once again "we" is used to mean the chumps who accept uncritically whatever is the fashionable view.

What Sivaram would like to happen: "He thinks government should increase its support for energy research and development, aiming at diverse pathways, applied at various stages of technology development, and targeting game-changing breakthroughs."
What really happens: a Democratic party fundraising bundler is rewarded with millions of dollars of subsidies for a solar start-up, which goes bankrupt.

+1. Its amazing how good Dems are at picking winners with other people's money.

"... possibly orbital solar power satellites." Which, perhaps, could also be used as a weapon. Which might enhance funding opportunities?

"Promoting solar energy also isn’t in the interest of regulated utilities. They fear a scenario where many users deploy solar power to detach from the energy grid, either wholly or in part."

Solar advocates think grid power should be available to them whenever their solar systems fail to deliver all the power they want, whenever they want it. BUT, really, it's somehow "unfair" that they should have to pay actual costs for this backup service.

And the more rabid promoters insist that not only should the utility be obligated by law to provide power whenever they wish to buy it, but also that the utility should pay retail price whenever a solar producer wishes to sell power- even if the utility has no immediate need for additional power when the producer wishes to sell.

Perhaps a few general principles could go a long way toward balancing the interests of utilities, solar producers, and consumers of electric power?

And, perhaps, government subsidies would be better spent on R&D than on deployment?

“… possibly orbital solar power satellites.” Which, perhaps, could also be used as a weapon. Which might enhance funding opportunities?"

The only effective transmission scheme would involve conversion to microwaves. So yes, orbital solar power satellites would be giant directed energy weapons.

>The only effective transmission scheme would involve conversion to microwaves. So yes, orbital solar power satellites would be giant directed energy weapons.

Hush. Those of us in the Military Industrial complex are trying to sneak this one past the greenies.

The energy per square meter is inherently pretty low. If it were higher, the prospects for space based solar power would be much better. So, it's not much of a weapon. There are laser power beaming technologies which would be more weaponizable, but efficiency suffers too much outside of specialized applications.

The military interest in it lies solely in it being one of the few ways to rapidly deploy a gigawatt of difficult-to-interrupt 24/7 power to whatever godforsaken location they might be sent on short notice.

For the rest of us, orbital solar power satellites are one of the few reasonable long-term alternatives to nuclear power. And probably the only way solar is going to be more than a niche solution.

"The energy per square meter is inherently pretty low."

The solar power is collected by solar cells. But you need to convert it to microwaves to transmit it to the ground efficiently. At that point you have a directed energy weapon. If you point any concentrated beam of energy in the 10's of megawatts range (let alone gigawatts) you are transferring a whole lot of energy to the target. Granted, you could intentionally disperse the beam and target it at a large receiver. Think several acres of wire netting on a bunch of poles with cows grazing under it.

But what country/group do you trust to not weaponize it?

Why bother? Once you're at the top of the gravity well, you can far more easily just rod anyone underneath you anyway. The station itself is probably a bigger hazard than the beam.

The wavelength determines how tight a beam you can get. With microwaves and their good transmission properties, you're inherently talking about antenna miles wide and power densities safe enough for birds to fly through and plants to grow under. You cannot get a tighter beam via some focusing error or antenna high-jacking. If I remember this correctly, the product of the diameters of the transmitting and receiving antennae must approximately equal the product of the range of transmission and the wavelength. There might be a constant in there.

While there has been some talk of much higher frequency microwave transmission, most proposals suggest beams in the 2-5 GHz range for water absorption reasons. The higher frequency designs might allow smaller focuses, but they'd have trouble with, among other things, rain.

There's no accident waiting to happen here. TallDave's right, too, that there are simpler ways to weaponize space if you want to do that.

"In short, the optimists are optimistic. Fine… now let’s get on with working through the problems."

Serious people have been working on this for decades. It's not like these problems just showed up in the last couple of years.

It's 92.96 million miles away to be exact. Which is the problem.

Look, solar has a bright future as a niche technology, but it won't ever be anything else until it gets a lot close to the Sun, probably hundreds of years from now. Too diffuse, too subject to being significantly blocked by atmospheric elements even when the Earth is facing in right direction, too bulky. These will continue to be true even if the cost per solar panel is zero and they never need maintenance and last for 1,000 years before dissolving into a pile of ecologically-friendly, 100% resuable parts.

Thanks for the book recommendation.

That's not really correct. Solar and wind are cost competitive now ignoring the costs of energy storage. The economics of energy storage for renewables have become the bottleneck, since that's the one part of the equation where costs haven't rapidly dropped.

Meaningless conditional metric. A hamster on a wheel is a cost-effective energy producer, if you ignore the problems keeping them focused.

Also, keep in mind, even if we get the magical low-mass, low-cost, low-environmental impact, high-efficiency batteries, the other power technologies benefit too, if not as much (every grid has to deal with peak conditions anyway). Often it's still going to make as much or more economic sense to charge the magic batteries using your other power sources during off-peak times, because of the fixed costs.

TD, you are ignoring the underlying economics. Wind power is already one of the cheapest marginal producers of electricity. It's more expensive than natural gas, but cheaper than coal. Furthermore, the costs of wind power have been steadily dropping (and the capacity factor has gone up) as the unit size has increased.

Furthermore, the cost of future wind from an existing wind turbine is very cheap. Whereas, a coal or natural gas plant still has to purchase fuel after it's capital costs are paid for, a wind turbine has no long term fuel costs. And maintenance on electric motors is lower than any boiler based technology. Once the capital costs are recovered, and existing wind turbine will need the turbine rebuilt every 20 years or so. The cost of a rebuilt turbine is obviously only the fraction of constructing the original site, installing a pole, adding a turbine, blades and a transmission system. So, one can reasonably assume that any large wind field will (barring unexpected changes) be producing as much or more power a century from now as it does today, for roughly 1/4th the current cost.

However, the intermittency of generation seriously increases it's effective cost. So ultimately the cost of energy storage, determines the long term viability of both wind and solar.

"Often it’s still going to make as much or more economic sense to charge the magic batteries using your other power sources during off-peak times, because of the fixed costs."

All those other sources use fuel and the marginal cost of fuel implies that wind will be the cheapest generation source.

Again, the hamster on a wheel is a cheap marginal producer too.

Wind can work great as a niche application in areas where the land has no conflicting uses and wind is steady (although for some reason it still needs billions in subsidies). Unfortunately there aren't a ton of those places nearby major sources of demand, and we're using up the best cases pretty quickly. And again, it's nearly useless without the magic batteries, because on the whole wind is even less predictable than sunlight.

"All those other sources use fuel and the marginal cost of fuel implies that wind will be the cheapest generation source."

No one really knows that, because it's not that clear how long the turbines will last (high winds are difficult to predict), or what future fuel prices will look like. Also, fuel is an insignificant portion of the price of nuclear power, and there's (at least) thousands of years' worth.

"Again, the hamster on a wheel is a cheap marginal producer too."

No, a hamster uses expensive fuel.

"Unfortunately there aren’t a ton of those places nearby major sources of demand, and we’re using up the best cases pretty quickly."

This isn't really true. As the wind turbines get taller, the areas that they can effectively use grow. But even at the current size, the US easily has enough land to add hundreds of thousands of additional turbines. Due to turbulence, you generally space them out at 50 acres per turbine, but the actual foot print is less than an acre. So, if doesn't even significantly impact the existing use of the land. The US has millions of acres of high wind prairie land.

"And again, it’s nearly useless without the magic batteries, because on the whole wind is even less predictable than sunlight."

Actually wind has a better capacity factor than solar. 40% for modern US wind fields versus 25% for solar. Granted, sunlight is more predictable, but it's still intermittent. And since you can't count on either as a 100% source, you have to have 100% backup capacity. The higher capacity factor wins out. Y

"No one really knows that, because it’s not that clear how long the turbines will last (high winds are difficult to predict)"

They're built pretty well.

"Wind turbines withstand tornado with nearly 300-mph winds"

"Also, fuel is an insignificant portion of the price of nuclear power, and there’s (at least) thousands of years’ worth."

As soon as we can build affordable nuclear power plants, I'm all for it. But currently wind power is a far cheaper solution.


""And again, it’s nearly useless without the magic batteries, because on the whole wind is even less predictable than sunlight.""

You don't need magic batteries. You can just use existing natural gas plants. If it helps, think of wind power as a fuel savings device for natural gas plants. When the wind picks up the natural gas plant goes into "low fuel usage mode".

When the cost of wind power drops below the marginal cost of natural gas, it will become the marginal generation source. It certainly won't need and shouldn't get subsidies.

"No, a hamster uses expensive fuel."

Who said anything about feeding them?

It's almost always going to be cheaper to just run the gas plant all the time, and not build the wind farms. Hence...

Going to make replacing the turbines much more difficult if the company making them went bankrupt a few years after the wind farm was built.

"The US has millions of acres of high wind prairie land."

And they're full of little prairie animals that could run on wheels, too. Neither of these is going to help much in supplying power to major urban areas, where the vast majority is used.

Note that we've already, even at today's tiny wind penetration, spent billions of dollars on transmission projects just for wind farms... in Texas alone.

They're a boondoggle outside a few place where the economics makes sense. People looking at the generating costs are not being realistic.

Except when they aren't.

"They’re built pretty well."

As soon as we can build affordable nuclear power plants

That's been true since the 1970s.

The primary subsidy for wind power has already started phasing out. It's 80% for 2017, 60% for 2018, 40% for 2019 and 0 after that.

Assuming that schedule doesn't change (with Congress, who knows), then we'll know within a few years if wind power can stand on its own or not.

There are a few places it can work pretty well... unfortunately much of subsidy is indirect, in the form of those expensive transmission projects, and it's an open question what politicians will do. Had there never been a subsidy, the less economic plans would have been shelved, which would have been better for everyone.

"unfortunately much of subsidy is indirect, in the form of those expensive transmission projects,"

That might be true in some cases, but it's not true in Texas. The new high voltage lines are being paid for via a $4 per month surcharge on users bills. That's not a subsidy, it's a user fee. It's purely an infrastructure upgrade to allow East Texas access to the cheap wind power in West Texas.

"Had there never been a subsidy, the less economic plans would have been shelved, which would have been better for everyone."

I'm not a fan of subsidies either and I'll be glad when (if) the Federal PTC is phased out. Personally, I think that wind power, at least, will still be economically viable. And that currently the subsidies are just fattening up the profits for GE and other politically connected corporations.

Thanks Alistair and Battery Boy for the input.

My 2 cents - chemical batteries will never work. Storage like super caps and old school pumping (locations are limited) will need to be the answer.

Locations do not seem to be limited if you start using old mines for pumped storage - see above concerning Bottrop.

I'm not convinced that their is sufficient volume in old mines to be a significant factor. Furthermore, I suspect that the water pumped into an old mine, might well require filtering. If you start pumping water into and out of an old heavy metal mine on a daily basis, you are going to have to deal with a lot of contaminated water.

>Is effective solar power further away than we had thought?

Certainly not further away than anyone whose been paying attention has thought.

If you're the kind of dope who hates oil and thinks electricity comes from it, than probably yes.

It is funny to contrast this to the robot discussion. Over there a "security robot" is defeated by bums with a tarp and barbecue sauce. People are like "technology rules!"

Here, gigawatts (literally billions of watts) of solar are deployed and people say "it will never work!"

Go figure.

I love how solar, wind and battery prices keep dropping and capacities keep rising, yet the conversation never changes on this board.

Tyler, and many people on this board seem to suddenly forget what learning by doing is when looking at something they instinctively distrust.

Looking forward to seeing the same crew and same arguments in ten years when wind and solar produce around ~40% of the electricity supply nationwide

I would be thrilled if wind and solar accounted for 40% of energy consumption (assuming energy use hasn’t fallen dramatically).

I do not think it is feasible in any reasonable time frame. The details have been discussed to death already. But the bottom line is intermittent vs steady. You need a base load of steady generation. Wind and solar could maybe help support that.

What makes me take the global warming crowd much less seriously is their lack of intense support for nuclear energy.

Right, because "global warming" is not a molecular interaction to solar radiation, it is a "crowd."

Depends on what "effective" means and how gullible "we" were about promised improvements.

Solar is great for isolated things that need relatively little power.

For baseload power, FFS, people, just go nuclear.

"Is effective solar power further away than we had thought?"

That's a good headline Tyler. You've managed to avoid Betteridge's law. I don't believe anyone authoritatively knows the actual answer to that question.

Not further away than I had thought, but do think eventually it will be there and we look bad at wind as a complete waste of money.

To replace fossil fuel electric generation with nuclear would require overcoming technical barriers at least as significant as those standing in the way of large-scale solar adoption. It remains possible that economic fusion is going to eventually deliver our golden age, but fission has not replaced coal and natural gas because the short-term benefits always end up getting eaten by the deadly combination of catastrophic tail risk and long-term costs.

"To replace fossil fuel electric generation with nuclear would require overcoming technical barriers at least as significant as those standing in the way of large-scale solar adoption."

France managed it without much trouble, and that was decades ago.

Only by having the government heavily manage the market. If your suggestion is the Federal government would centrally manage electricity policy, I'm pretty open to thinking about that idea, but I think a lot of stakeholders and the entire Republican party is going to be skeptical.

That's not a "technical barrier."

Also, utilities are already very heavily regulated.

And I'm not arguing the US should convert to nuclear power anytime soon (although it is likely we will eventually want to), just that there aren't any serious "technical barriers" to doing so.

The first disquieting sign is that solar companies are spending only about 1 percent of their revenue on research and development, well below average for a potentially major industry. You might think that’s because things are going so great, but some major solar users may have already maxed out their technology. According to Sivaram’s estimates, four of the five most significant country users — Italy, Greece, Germany and Spain — have already seen solar energy flatten out in the range of 5 percent to 10 percent of total energy use. The fifth country, Japan, is only at 5 percent.

If the solar R&D is too low, then what accounts for the well documented decline in the cost of solar power modules over the last few decades (for example Is this just caused by economies of scale? If so R&D investment might be premature. If you can cut costs by 50% or more by achieving economies of scale, it might make sense to go there before you start trying to use R&D to make major new breakthroughs.

As an R&D engineer in the PV world, I just wanted to add my two cents here. It is true that overall R&D spending is quite low in the PV industry. I believe the reason for this is that Chinese PV manufacturers comprise over 50% of the market, and spend relatively little on R&D due to their already low margins. This chart ( shows the R&D spending over time of the top PV R&D spenders. The top two companies (by far) are two US companies, First Solar and SunPower, which require significant R&D spending to maintain a technological lead over their lower priced Chinese competitors. These two companies spend roughly 3-5% of their annual revenue on R&D which is close to the average R&D spending over all industries. An imperfect analogy would be with cell phones, with higher reliability, higher cost, higher margin products from American companies like Apple competing with lower reliability, lower cost, lower margin products from Xiaomi et al. However, price is not the end-all-be-all in cell phones as it is in solar panels, and thus Xiaomi has not dominated market share the way Chinese PV has.

Regarding future progress, the top Chinese solar companies are in the midst of a technological shift (to n-PERC, for the curious) which should increase their efficiency, though not to the level of SunPower. They also continue to scale; these two trends in tandem will continue to drive down costs. First Solar and SunPower (which have differentiated technologies, both from each other, and from the Chinese) are also continuing downward trends in costs through their aforementioned R&D expenditures. Module costs are not the whole story though, and "soft costs" (customer acquisition, permitting, development, etc.) are starting to become bigger and bigger pieces of the cost pie. This, however, is not my area of expertise, so I can't comment intelligently on the trends in this area, but I know that it is an area of increasing focus in the industry.

Do you have any thoughts about the China dumping issue? Are they genuinely selling below manufacturing costs? Using state subsidized loans? Or just cranking out the product really cheaply?

Solar PV is good for off-grid and as a low-carbon daytime peaker in sunny places that can use that. 10x cheaper storage might change that a bit. Penetration of 5%-10% is about what can be expected and that's still a pretty big business. It's silly not to think of energy production as a regionally optimized portfolio where you will rarely have 100% of anything.

I'm interested in knowing if Tyler has solar panels on his house.

He's a smart guy. I'm sure he makes a decent living.

If he has solar panels - why?

Or why not?

Liddell is an aging black coal power plant in Australia scheduled to shut down in 2022. Australia is a low cost producer of coal and the largest exporter of it in the world. The current federal government is in favor of coal so they wanted to spend hundreds of millions of dollars to extend it's life by 5 years. The owners of the power plant, the multi-billion dollar company AGL, which has shown itself quite happy to pollute air and contribute to greenhouse gas emissions in the past, refused because it would be cheaper to build solar power and wind power capacity plus dispatchable gas peak generators. An outline of AGL's plan can be found here:

New coal power stations cannot compete economically with the cost of a combination of distributed and utility scale solar, wind, and peak gas turbines with possibly some gas reciprocators thrown in, so Australia will never build another new coal power plant.

If coal has no future in Australia, it has no future anywhere.

That's an interesting article. It doesn't look like the proposed replacements would work however. Trying to replace a 2 GW coal plant with 1.6 GW of renewable + 0.75 GW of Peaking plants + 0.1 GW of upgrade to another plant doesn't add up. The math doesn't work.

I'm a proponent of renewables where they make economic sense, but you can't just replace a baseload power plant with the same amount of renewables.

There will be periods when the renewables 0 out, and indeed, you would expect that the average day output for 1.6GW of renewables is around 0.7GW. So it seems like the 'peaking' plants will run nearly all the time and on days when renewables aren't producing you'll have to draw power from other parts of the grid.

Well, that is just what AGL says they are building to replace Liddell. Their goal is to make money and they are under no obligation to build what the grid actually needs. They will happily blackout Australia if it somehow makes them money. But there are other developments going on in the grid and they will be taking them into account and they will adjust their plans as time goes by.

Looking at the list in the article I see it adds up to 1.25 gigawatts of dispatchable supply and effectively dispatchable reduction in demand. If the renewable capacity is around 10% firm that's another 160 MW bringing up to around 1.41 gigawatts. That 's not too bad compared to an unreliable coal plant that can't manage to produce its 2 gigawatt original rating.

Solar will continue to grow and there will continue to be innovation in solar cell technology.
For example:
Ultra-thin solar cells can easily bend around a pencil
This is early days for solar. Sort of just past the Wright brothers, well before the jet.

Also, solar electricity is already at $0.01 per kWH at the equator and this may lead to big changes in the aluminum, fertilizer, concrete and other energy intensive industries. Those big deserts in the middle east, particularly in Saudi Arabia, may become the home to these industries.

OK some thoughts, some of which might have been covered, others maybe not?

1. R&D % is not as insightful as it first seems. First PV costs have been dropping dramatically over time. That means makers are either innovating without formal R&D or they are quickly dropping down an economies of scale curve. If you have deep savings that can be accomplished by economies of scale, does it make sense to invest in R&D to maybe leap to the next level or invest in increasing capacity to get to the lowest possible cost? Second you have the Apple phone effect. Apple spends a bundle on making the next iphone the greatest possible. Lesser name cell phone manufacturers are not trying to out R&D Samsung, Apple or Google. They will try to chase market share by achieving the lowest possible cost.

2. Yes 100% solar based grid seems unobtainable but that's a strange metric. Today the grid is hardly 100% of any one thing. What in life is supplied by 100% one thing? Is there one food we expect to feed 100% of us? One vehicle that supplies 100% of transportation? One plane for 100% of aviation? One type of investment for 100% of your retirement portfolio? Why is solar tarred as a failure because it is unlikely to achieve 100% of the grid but, say, no one knocks Facebook for being unlikely to ever get 100% of the social media market?

3. Curious about 'artificial photosynthesis'. If surplus solar power could be turned into hydrogen, gas or even liquid gas you have solved the 'storage problem'.

4. Budgeting for market change and innovation: Ok one thing to think about is if solar ramped up one possibility to me seems to be that you'd see some huge swings in electricity prices....possibly dropping to $0 or even negative at peak production times or going up when demand is high and it's dark out. What this is doing is dramatically increasing the possibility of arbitrage profit for anyone who can either buy power when it is cheap, store it and then sell it when it is dear, or transfer demand from peak to slack periods. In fact those are essentially the same thing since anyone running a 'battery farm' would essentially demand more electricity whenever a solar surge causes a glut. Rather than just looking at the prospects for huge innovations in battery technology, we should consider a market response could be driven by something as simple as a appliance apps that adjust their operation by looking at short term electric prices. If the market is there then it shouldn't be a given that we need Star Trek level technology breakthrus to adapt to it.

5. Knock on benefits, what exactly would happen if every roof had a solar panel on it? If we start making things like roof tile/shingles solar systems, well almost all buildings need to have roofs, no? If the solar panel does a decent job as a roof or paint and the cost difference between that and 'dumb shingles/paint' is small then that side benefit should be counted as well.

There is hope for you academics.

Bhojpuri Songs Free Download services

Comments for this post are closed