# The Holy Grail of Battery Storage

A recent Telegraph article claims that storage battery technology is now advancing so fast that “we may never again need to build 20th Century power plants in this country, let alone a nuclear white elephant such as Hinkley Point” and that the “Holy Grail of energy policy” that will make this solution economically feasible – a storage battery cost of \$100/kWh – will be reached in “relatively short order”. This brief post shines the cold light of reality on these claims by calculating battery storage costs based on the storage requirements for specific cases estimated in previous Energy Matters posts. It is found that installing enough battery storage to convert intermittent wind/solar generation into long-term baseload generation increases total capital costs generally by factors of three or more for wind and by factors of ten or more for solar, even at \$100/kWh. Clearly the Holy Grail of energy policy is still a long way off.

First a simple calculation. \$100/kWh = \$100,000/MWh = \$100 million/GWh = \$100 billion/TWh. If everyone is happy with this we can proceed. (Note that all the costs listed in this post are in US dollars unless otherwise specified).

In the Is large-scale energy storage dead? post I presented this graph:

The procedures used to estimate these storage requirements are described in these posts:

Multiplying the storage capacities shown in the Figure by \$100 billion/TWh gives the following battery installation costs. Wind and solar installed costs (both estimated at \$2,000/kW) are from IRENA :

Battery storage needed to convert Germany’s 2013 solar generation to baseload: \$800 billion, about 13 times the \$66 billion cost of installing the ~33GW of solar capacity involved.

Battery storage needed to convert solar generation equal to a year of Hinkley nuclear generation to baseload: \$700 billion, about 28 times the ~\$25 billion cost of the Hinkley  plant.

Battery storage needed to convert solar + wind generation equal to a year of Hinkley nuclear generation to baseload: \$350 billion, about 14 times the cost of the Hinkley nuclear plant.

Battery storage required to convert one month of UK wind generation to baseload: up to \$500 billion, over twice the \$200 billion cost of the ~100GW of wind capacity involved. (Note 1: storage requirements for a complete year would likely be significantly higher. Note 2: the lower-storage options discussed in the “estimating storage requirements” post are achieved by increasing wind capacity and curtailing large amounts of wind power.)

I added a small project– Gorona Del Viento – to round the estimates off. During its first year of operation GdV generated only about half the wind energy needed to fill El Hierro annual demand, but had it generated 100% of it then 10GWh of storage would have been required to store the wind surpluses for re-use in windless periods. The cost of installing this much battery storage is \$1 billion, approximately ten times the €82 million euro project capital cost.

And how good are my storage estimates? Well, the late Sir David Mackay, at the time DECC’s Chief Scientific Adviser,  confirmed some of them in a comment on the “estimating storage requirements” thread:

Your calculations agree with my back-of-envelope estimates. In SEWTHA Ch 26 I said “imagine we had 33 GW of wind capacity, delivering on average 10 GW”; I reckoned that ballpark of 1000 GWh of storage would be needed ….

Mackay’s estimate gives a ballpark battery cost of \$100 billion, not quite twice the \$66 billion cost of installing his 33MW of wind power but again well in excess of it.

Clearly large-scale battery storage will remain uneconomic even at the Holy Grail price of \$100/kWh. So why do battery companies, research institutes and greens claim the opposite? Because they assume that the intermittency problem can be solved simply by installing enough storage to balance daily load fluctuations. A large amount of storage isn’t necessary to do this, and \$100/kWh batteries might indeed be able to supply it without breaking the bank. But they ignore the much larger amounts of storage that are needed to keep the electricity coming during extended windless periods and/or to flatten out seasonal variations in solar output. Why? I see two possible explanations. First, they are being carried along in a wave of visionary enthusiasm and haven’t recognized it as a problem; second, they know about it but don’t want to tell anyone because it might spell the death of large-scale storage battery research, and ultimately maybe the death of intermittent renewables too. I’ll let the readership make up their minds as to which it is.

This entry was posted in Energy, Political commentary and tagged , , , . Bookmark the permalink.

### 120 Responses to The Holy Grail of Battery Storage

1. roamer12345 says:

But I don’t think anyone, at least in the US is serious about a 100% renewable grid. Natural gas looks like a pretty good battery for the grid for the foreseeable future. \$100/kWhr will not pave the way to a renewable grid, but it starts to make electric transport look more viable.

• Willem Post says:

roamer12345,

Using Roger’s method, why not calculate the cost of energy storage systems for just 50% renewables to cover the much larger amounts of storage needed to meet demand during extended windless periods (which can occur at any time and at random) and low sunlight periods (which occur mostly in winter, due to snow, ice, fog, etc.), and to flatten out seasonal variations in solar output.

With many PV systems on a distribution grid, the output variation during daylight hours (variable cloudiness) would disturb those grids, unless they have battery damping/storage, as is being implemented in southern Germany and southern California. The cost of that storage is “socialized”.

In New England, about 75% of the hours of the year, PV energy would be minimal or zero.

The production ratio of MONTHLY maximum summer/MONTHLY minimum winter is about 4, based on data from numerous existing PV systems in NE.

The DAILY ratio is about 25.

In New England, about 40% of the hours of the year, wind energy would be minimal or zero. In addition, the output continually varies 24/7/365, just as the wind.

Those numbers are even worse at higher latitudes, such as in Germany, etc.

BTW, do not forget the about 20% A to Z loss of energy, due to the various losses of battery storage systems, including parasitic losses, and before and after transformer losses (if higher voltages are involved), and before and after inverter losses.

• Willem Post says:

Addition to above comment:

See below article regarding battery storage, based on real world numbers:

http://www.theenergycollective.com/willem-post/2308156/economics-of-batteries-for-stabilizing-and-storage-on-distribution-grids

• jacobress says:

“In New England, about 75% of the hours of the year, PV energy would be minimal or zero.”
Wrong.
In Israel, 32 deg Lat, PV produces energy about 1740 hours a year, which is 20%.
I guess in New England (like in Germany) PV production would be zero about 88% of the hours of the year.

• Willem Post says:

jacobress,

I guess in New England (like in Germany) PV production would be zero about 88% of the hours of the year.

How did you arrive at your guess?

• Alex says:

A capacity factor of 12% does not mean zero production for 88% of the time.

“Zero” production is for around 50% of the time. Minimal production (say less than 100W from a 5KW system), perhaps 2/3 of the time.

• jacobress says:

Well, zero or close to zero.

• Euan Mearns says:

This is part of the hypocrisy of renewable fantasy. It started out as distributed generation so we could do away with centralised units and a vast distribution network. Now it has morphed into vast centralised renewables units in the wrong place connected to consumers by a gigantic distribution net.

• robertok06 says:

I totally agree on this Euan…. the hypocrisy of the greens is astonishing.

• jim says:

I think it is more a matter of complete ignorance of basic science and logic on top of having no concept of reality plus a gullibility for every green Charlton that comes along.

• Ikemeister says:

roamer12345 you say: “I don’t think anyone, at least in the US is serious about a 100% renewable grid.”
Do you speak for Mark Z Jacobson then? He continues to claim it is quite feasible.

2. roamer12345 says:

Point well made regarding economic feasibility of a purely renewable energy grid using \$100/kwHr batter tech. I don’t however see why higher levels of renewable integration with natural gas CHP as effective battery won’t sans nuclear.

• Euan Mearns says:

Nat gas is not a battery. It is a fossil fuel energy store. Be careful to not twist the terminology any further to suit you ends. Renewables have locked society into FF dependency. While nuclear may have largely freed us.

• RDG says:

“Renewables have locked society into FF dependency.”

I would argue that the dependency (which nuclear cannot free ourselves from) is the fraudulent economy of pumping up real estate values on hot air by letting the industrial base fall apart. The hardcore renewable fanatics seem to be more interested in saving the real estate bubble in their Musk-like dreamworld of further destroying the industrial base and powering houses and vehicles with rooftop solar and batteries that are “cheap”.

How do you keep propping up real estate if you have to shove a ton of dough into a new nuclear build?

3. Peter Lang says:

An interesting reality check would to project battery storage costs in future using the learning rate since batteries were first made in 1800. For example the learning rate for coal fired electricity generation is 12% since 1900 http://arxiv.org/pdf/1001.0605.pdf (Figure 6, bottom). For nuclear it was 23% to 35% for six countries (excluding India) from 1954 to about 1970, and 33% in South Korea since 1973. https://judithcurry.com/2016/03/13/nuclear-power-learning-rates-policy-implications/ We can project that nuclear power could have been 1/10th of current cost in 2015 if the pre-1970s learning rates had continued.

I suggest the root-cause of the disruption (see Figure 2) was the growing concerns about, and the loss of public confidence in, the safety of nuclear power; these were largely a consequence of the scaremongering by the anti-nuclear power protest movement, which became influential in the mid 1960s. The anti-nukes have a lot to answer for, including an estimated 4.5-9 million deaths that could have been avoided and an additional 75-170 Gt CO2 emitted.

The technology learning rates over history are a good way to get a reality check on what the rate of cost reduction is likely to be in future.

Suggestion: an interesting future post?

• jacobress says:

“We can project that nuclear power could have been 1/10th of current cost in 2015 if the pre-1970s learning rates had continued.”

Why stop in 2015? At the same “learning rates” the cost will turn negative around 2025!

• Peter Lang says:

Learning rate is fractional cost reduction per doubling of capacity or output. It never goes negative. it reaches zero coat and infinity doublings.

• Peter Lang says:

If the global deployment rate that prevailed up to 1976 continued from 2020 to 240, and if the learning rates that prevailed up to 1970 continued from 2020 to 2040, the cost would about halve.

The point of the exercise is to illustrate the enormous damage the irrational anti-nukes, like you, have done to human wellbeing.

4. Send the author a message with this link.
ambrose.evans-pritchard@telegraph.co.uk

• Joe Public says:

But don’t expect a response.

• robertok06 says:

Writing to him would be a waste of time… he knows nothing about technology and science.

• Gaznotprom says:

When I read the article and another one on Wind farms, thought Ambrose Prichard…. had swallowed & digested, overnight a huge marketing presentation from the ‘renewables’ industry, feeling queasy bringing it up in the morning…

Again, good factual analysis Mr Andrews, thank you!

5. Joe Clarkson says:

CSP (steam over air) with sensible heat thermal storage in the form of gravity contained pebble beds (not molten salt or molten glass) at a cost of about \$3.00 per kwh (electric) could actually provide seasonal storage. Why all the fuss over storing electrons?

• Greg Kaan says:

Can you provide a source for that costing? It seems extraordinarily cheap and at that price, pursuit of grid scale batteries would indeed be foolish.

• Leo Smith says:

Of course he cant. Seasonal storage means a very large heat bank indeed. And what you get out is low grade heat, not electricity.

Yes, an Olympic sized pool of insulated hot water will heat your house in the winter, before its too cold to be useful, but that’s all it will do.

And using electricity to heat is not really a great idea in the first place.

6. John F. Hultquist says:

Until intermittency is 100% solved some other technology, such as gas, has to be in place for the times needed. Do the folks making claims for storage not have refrigerators and freezers, or other products and services that need electricity? If the backup is used only a little, who pays how much to keep it there? Taxpayers will end up “owning” the entire industry.
Maybe that’s the plan.

7. sod says:

This is a very pessimistic outlook and it is really easy to see, why this comes to a completely different result than for example the Telegraph article (among all sources, by the way!!!).

The calculation that shows nothing but problems works like this: How much would it cost to do seasonal storage with batteries at costs from today (or the \$100 number from the Telegraph article). This ends up with extreme numbers.

The positive outlook goes like this: Battery prices are dropping and there is huge potential for massive improvements (about the opposite of those possible in a coal plant, by the way). Each step will have immediate benefits: small amount of grid storage will already help to smooth all kind of problems. Home storage will provide a level of independence. Small amounts added to solar PV will already bridge the gap to the evening peak.Electric cars will provide both, a different form of transportation AND storage options. All of these developments will drive down prices and will make the next step so much easier.

a positive view also looks at the transformation done to batteries and their use today. These mostly come from cell phones and mobile computers. The pressure for improvements is incredibly high (again, in total contradiction to improvements of coal plants. Basically nobody cared about those at all over decades!).
Another big change is, that we finally have a chain of batteries from small to huge (while in the past we were trying to jump from the torch light to the electric car)! Starting with cell phones, all sort of garden tools, electric bikes (very popular today, and giving people much needed confidence in battery technology!), electric motorbikes, electric cars with a pretty big range, power walls, battery UPS for data centres and finally small attempts of grid scale storage.

So battery technology will be a huge part in “replacing” for example Hinkley plant in the late 2020s. But it will not be in the form of huge grid size batteries, storing electricity for a week of no wind in winter (this obviously does not make any sense)..

• Greg Kaan says:

The pressure for improvements is incredibly high (again, in total contradiction to improvements of coal plants. Basically nobody cared about those at all over decades!).

This statement could not be any further from the truth. The search for improved batteries has been continual ever since the galvanic cell was first invented.

If you want an example of where battery performance (in every regard) is critical, you don’t have to look any further than the field of submarines. A breakthrough in battery technology would transform the performance of all non-nuclear submarines. Denying that well funded research into batteries never took place until recently ignores the entire history of submarine warfare.

And this is just one example.

• Leo Smith says:

Exactly. Its more airy fairy handwavey greenshill nonsense

intermittent renewables are way more expensive than nuclear without storage. adding storage will simply make them even worse.

• sod says:

“If you want an example of where battery performance (in every regard) is critical, you don’t have to look any further than the field of submarines.”

Thanks for the serious reply.

I doubt that submarine technology is applying any kind of pressure on a market like batteries.

In total contrast, cell phones and electric cars are such a driving force.

It is obvious, that electric cars needed to increase their range by a factor of about 10. And this is obviously being done.

The submarine market simply could never have induced such a change.

I notice the changes every day. my cell phone is loading fast. My electric drill is always loaded when i need it. after some problems with the first battery on a pedelec, it looks like everyone around me is pretty happy with the one they have now.

• robertok06 says:

“It is obvious, that electric cars needed to increase their range by a factor of about 10. And this is obviously being done.”

Being done by whom? Which manufacturer?
Tesla’s car fleet has not improved at all its mileage, nor the distance it can run with a fully loaded battery.
A couple of weeks ago I’ve watched on TV a guy going roundtrip from Paris to Nice and back, with a Tesla… average speed taking into account the biblical recharging times?… 80 km/h… like in the 60’s, when there were no highways.

You clearly live in fantasy-land.

• Leo Smith says:

how much are they paying the green shills and astroturfers to hang out here?

• robertok06 says:

Don’t be silly… they spread the verb for free… to save mankind.

• Dave Ward says:

“How much are they paying the green shills and astroturfers to hang out here?”

Rather less than at Pierre Gosselin’s “NoTricksZone” blog, where “sod” displays his stupidity in every thread.

8. clivebest says:

We already have batteries that can store 1KWh for less than \$100 – lead-acid car batteries. A 12V car battery rated at 100 amp hours can provide 5 amps for 20 hours. So the total energy stored by the battery = power x time or 12x5x20 = 1.2 KWh !

How much would it cost to power the UK with such batteries over a 24 hour period without wind in December ? The UK needs about 0.8 TWh of electrical energy per day. This could be provided by about 7 billion car batteries, each apparently costing around £50. So this national energy store would cost just £350 billion !

The only problem is the amount of lead that you would need. This would weigh in at about 91 million tons, and currently the known reserves of lead are just 80 million tons. So unfortunately the prices would rise exponentially as reserves are used up.

The same argument can be made for Lithium.

• Euan Mearns says:

Clive, it was a question I had for Roger that maybe you could answer. What does 7 billion car batteries look like in terms of volume and mass and what is involved in their manufacture in terms of raw materials – squeaky green lead and sulphuric acid?

And how long does the battery array last? Our baseline is always Gen III nuclear with projected minimum life of 60 years. How many times do the renewables machines and battery back up need to be renewed in that time frame?

• Another thing to note is that batteries will need a small but incoming electrical supply to keep them charged.

• Beamspot says:
• clivebest says:

I calculate the batteries would need 91 million cubic meters of space. You would also need space for cooling so realistically that is 300 million m^3. That is about the equivalent of 500 times the Albert Hall. I suppose you could distribute these in 500 different Albert Hall size locations.

Each battery probably has a lifetime of say 5 to 10 years so multiply the cost by about 8 to compare with Hinkley C.

• robertok06 says:

“Each battery probably has a lifetime of say 5 to 10 years so multiply the cost by about 8 to compare with Hinkley C.”

Your calculation assumes that the 7 billion car batteries are charged/dicharge at 100% every day… while car batteries do not take well at all a large depth of discharge, normally in a car the battery is kept close to 100% charge by the alternator.
If you assume a 50% DoD then your estimate doubles, and the costs, of course, do the same.

• Dave Ward says:

“Each battery probably has a lifetime of say 5 to 10 years”

Extremely unlikely for a “car” battery!

If you assume a 50% DoD then a typical car battery might give you 200 charge/discharge cycles. A proper “Deep Cycle” lead acid battery could provide around 1,500 cycles. But deep cycle batteries don’t provide or accept high currents as well as car type “starter” batteries, so you will need larger capacities to handle the considerable loads involved in daily peak levelling.

This is what the greens simply don’t understand – you can take a jerry can and fill it up with petrol or diesel in seconds. Then empty it just as quickly. But on top of that you can do this pretty well indefinitely. NO battery offers anything remotely comparable.

For more details of various performance parameters here is a link to one manufacturer’s top of the line batteries (and please note that I have NO connection with the company):

• Lets put it another way. Say we have a 4kWp solar on a roof in the UK (case is just for summer). This will give us on average around 9kWh per day (assuming no bad days*).

Around 70%+ of this is exported to the grid so lets say we want to keep this and use it for our use. That is about 6.3kWh to be stored, about the size of a power wall (with no headroom or losses, ignoring the W limit). Say 5 million of UK homes go for this.

Lets say 150g of Li per kWh (double theoretical, no idea here). Call it 5,000 tons of Li for the 5 million UK homes to load shift or just under 1% worldwide production assuming correct grade and no losses.

Assuming 5000 cycles and degradation over time, Tesla’s 10 year warranty seems to be pushing it a bit.

So the UK alone would be putting a significant dent into world wide production every 10 years. Clearly if going down this route, we need to load shift before using batteries.

*http://www.cambridge-solar.co.uk/solar-pv-cambridge/

• robertok06 says:

“Around 70%+ of this is exported to the grid so lets say we want to keep this and use it for our use. That is about 6.3kWh to be stored, about the size of a power wall ”

OK. So you’ve done this on the first day with lots of sunshine… your power wall is full… the next day is sunny again… what do you do with the 70%/6.3 kWh? Buy another powerwall?

Explain.

• donoughshanahan says:

I was only using the Powerwall example to look at lithium use and I see fairly significant problems. Thus load shift is required before batteries are used.

In any case 4kWp solar would produce (maybe) enough to cover the average daily electricity use of a house, just not at the right time of day, in summer.

The assumption is that by using a Powerwall you can load shift your generation away from midday to other use times. I am not challenging this here, just looking at Li.

• robertok06 says:

7 billion car batteries distributed around all households (25 million of them in the UK?) would mean 7000/25=280/household… this gives a good idea of the feat.
Also note that the magical quinone battery mentioned over and over and over again, by the Harvard scientist would need MORE space, as the flow batteries have a lower energy density as compared to those used in cars.
It’s a mission impossible even at 50\$/kWh.

• Thinkstoomuch says:

Minor nits.

Not all lead acid batteries are alike. using a car battery for storage is the worst way to go, IMO. Might as well use a quarter horse to pull a plow except the horse will be far better suited to the plow than a car battery to baseload energy storage. It has lots of amps for a few seconds. Ask my defunct motorcycle battery that is waiting on replacement after a single oops deep discharge.

My brother’s deep cycle 390 AH (20 hours discharge cap.) flooded lead acid 6 V batteries go for ~\$300 retail which is over 2 kWH. Realistically less than 1/2 of the 2 kWH can be used before charging. Further discharge shortens battery life significantly. Leading to the generator he uses to supplement the solar.

So for 1 KWH would still be \$3 for deep cycle flooded lead acid. Complete with maintenance costs to keep the water topped off. An absorbed glass mat type would be \$471. But no maintenance costs.

Granted there may be economies of scale if you are buying 1000’s but …

Like I said minor nit and a mostly useless comment.

Have fun,
T2M

Looked up prices for fun

For AGM

http://www.wholesalesolar.com/9960105/crown/batteries/crown-6crv390-390ah-6v-l16-battery

For Lead acid they say 430 amp hour
http://www.wholesalesolar.com/9960100/crown/batteries/crown-cr430-6v-flooded-l16-battery

Also gives sizing if you are interested.

• gah789 says:

Batteries don’t work like this – especially cheap lead-acid ones. Even the best deep cycle AGM or gel batteries cannot be discharged below 40-50% on a regular basis and they cost double a simple car lead-acid battery. In practice the cost is something like £150-180 per kWh of usable storage. With regular use you are lucky to get a battery life of 5 years, so multiply by 6-12 to get the gross equivalent of power plants with a minimum life of 30 years or a maximum of 60 years.

All this comes from the daily experience of running off-grid wind & PV systems for wireless broadband infrastructure, so we know how these technologies work in practice! Lithium-ion batteries are even less attractive for the foreseeable future.

• Alex says:

Lead acid batteries are not designed for deep discharge, and if used in this mode, only kast about 500 cycles. Although they have a low capital cost per KWh, they are more expensive than Li-ion in terms of usage cost (as Li-ion should be good for 5,000 cycles).

WE might be able to construct some weird hybrid sytem where lead acids are used for seasonal storage, and li-ion for daily storage.

As robertok said, 280 batteries per household. A double garage would do the job.

More likely would be flow batteries made from re-purposed oil super tankers, moored next to the giant wind farms in the North Sea.

To rephrase what you you say, it’s “mission impossible even at \$50 billion / TWh”.

• It’s interesting to note that arguably the world’s two most successful renewable energy projects – Eigg and King Island – both use lead-acid batteries for energy storage. King Island is reported to have burned out a bank of vanadium batteries before going the lead-acid route.

9. “I see two possible explanations. First, they are being carried along in a wave of visionary enthusiasm and haven’t recognized it as a problem; second, they know about it but don’t want to tell anyone because it might spell the death of large-scale storage battery research, and ultimately maybe the death of intermittent renewables too. I’ll let the readership make up their minds as to which it is.”

Having worked for three years in such an institution, I would say it is the former.

While the engineers and scientists there were working on solar or wind, they were focused on solving the problems related to production of the product and fining more efficient routes.

They would have no view on grid integration issues and no tools to solve that in the same way that a chemist making a drug in the lab (my brief stint at Merck) usually has very little scope on human trials.

However the final companies (and governments) who are selling the product, should have a view but then they leave it up to the grid (as do all energy generators). What has not been realized by the grids is integrating non baseload is far more problematic than base load, and thus they has seemingly not run the analysis.

The main point though is if you get sold something that is not quite right, you are culpable to some extent as well.

10. gweberbv says:

100 bucks/kWh is indeed a ‘holy grail’ – for the breakthrough of electric cars. But of course not for (utility-scale) stationary storage systems to provide electricity supply.
Probably the Telegraph article mixed things up.

11. Beamspot says:

Mmmm, let me rise a simple and stupid question. Life expectancy?

I mean, if they last 100 cycles, they are expensive. If they last 1M cycles, well, they may be cheap after all, but only need few years to pay back.

Not only cycles are the issues. Temperature (Arrhenius) also has something to say, not only to life, but also to power, efficiency, etc.

Round trip efficiency, net, including energy used by pumps, compressors, cooling/heating and such may be also relevant.

ESOI calculation would be much more relevant that this ‘advertisment report’ that is focused to kill Hinkley Point and keep green dreams alive.

I’ve read (in the previous post) really wild claims (460 years MTBF, while the best in semiconductor is about 100.000 hours, or about 11 years and half) that are more or less in the same line, and without any data backing (except, perhaps some early research wort at a lab under really ideal conditions, totally out of the real world).

Sounds like the propaganda machine is picking up steam.

Another point not explained is the size/weight of those options. Not really relevant for static grid backup, but parmount for any transportation issues. Last time I’ve read about flow batteries, they looked very elefantiastic regaring size, power, weight and such, reducing their application to static devices for peak load shaving or daily balancing, and that with some power limitation (as well as space).

And our economy is heavily dependent on FF transport technologies. Why nobody talks again about moving bulk freight to railroads/electric trains instead?

EROEI estimations?

• RegGuheert says:

“I’ve read (in the previous post) really wild claims (460 years MTBF, while the best in semiconductor is about 100.000 hours, or about 11 years and half) that are more or less in the same line, and without any data backing (except, perhaps some early research wort at a lab under really ideal conditions, totally out of the real world).”

Still making nonsense statements?

The 460 years MTBF is measure, on-roof performance of fourth-generation Enphase M215 microinverters:

MTBF of modern power mosfets at room temperature is over 1 BILLION hours (about 13,000 years):

https://www.fairchildsemi.com/product-technology/extended-temp-mosfets/

• Beamspot says:

Certainly there had been an engineer sitting 460 years to check that they didn’t broke. For sure. I’m doing nonsense statements. You are right.

And for sure those inverters don’t have any of those?

https://industrial.panasonic.com/ww/products/capacitors/polymer-capacitors/hybrid-aluminum

4000h.

BTW, the link you provide didn’t state any time for life, and those that I read, give on the shelf at cold temperatures, no temperature swings and such, not working at the rated temperature etc.

Like the LED’s used in traffic lights, with a MTBF of 100.000 h, where more than half never lasted 5 years, and they are on less than 50% of the time.

The conditions to reach the MTBF are quite strict… and really hard to achieve in real life. After ten years, when your inverter broke, try to sue the manufacturer (if it is still alive, like happened with many PV manufacturers), and check what will happen, what will they say.

Ah, and I beg your pardon for asserting data taken by myself when I was an R+D engineer at an Hybrid and Electric Vehicle dept in a well known european automotive transnational. Clearly unfunded assumptions that showed me that the datasheets released by manufacturers are, ehm, tricky (not to say plainly a fraud).

Lack of information and playing with words is the trick. Say that you canwithstand two charge discharge cycles per day 7305 cycles wihtout stating at which % of DoD clearly doesn’t qualify in front of a jury for 7305 cycles at >95% of DoD.

But in another part of the same datasheet it states that they can withstan >95% of DoD cycles, but not how many.

Again, it didn’t qualify in front of a jury for 7305 cycles at 95%DoD. It only means that they assure that, brand new, you can obtain >95% of the spec’ed capacity, and that’s all.

Been there, done that.

• Beamspot says:

I found the life expectancy of the fairchild transistors.

Funny.

They give a value stimation on a ‘real test run at stressing temperature’ of 150ºC during 2250000 hours, 257.13 years approx.

Since Fairchild is about 70 years old, I guess how thet measured that.

Another evident lie.

• robertok06 says:

One simply takes 1000 of them and let them go for 1 year… if after that there’s not a single one in fault they have already 8.76 million hours… let them go and record they failures and one gets the distribution… from which a MTBF can be calculated.
There must be some ISO rule/code which clearly states how the MTBF is calculated.

12. Jim Brough says:

When intermittent solar can provide electricity to run the Sydney Rails system for 22 hours every day I’ll be converted to renewables.
If renewables are meant to save the earth from uncontrollable global warming caused by CO2 emissions, why do supporters of renewables ignore nuclear electricity which has the lowest CO2 emissions per kWh.
To add battery storage to cope with solar and wind we will have to dig more mines and refine the materials at the cost of more CO2 emissions.
That way is lunacy. Gulliver’s Travels over 400 years ago had this to say about a university

The first man I saw was of a meagre aspect, with sooty hands and face, his hair and beard long, ragged, and singed in several places. His clothes, shirt, and skin, were all of the same colour. He has been eight years upon a project for extracting sunbeams out of cucumbers, which were to be put in phials hermetically sealed, and let out to warm the air in raw inclement summers. He told me, he did not doubt, that, in eight years more, he should be able to supply the governor’s gardens with sunshine, at a reasonable rate: but he complained that his stock was low, and entreated me “to give him something as an encouragement to ingenuity, especially since this had been a very dear season for cucumbers.” I made him a small present, for my lord had furnished me with money on purpose, because he knew their practice of begging from all who go to see them. (III.v)

13. Robert says:

Can’t we adopt the approach of replacing nuclear to a level greater than we need and selling spare base load capacity to Europe at inflated prices when there is some kind of interruption to their Russian natural gas supply or a month of heavy cloud

We can employ engineers and have skills and ‘stuff’. A few new reactors does sound a lot more neat a solution than millions of batteries everywhere

I think of it like memory for the computer. Do I want a distribution of dozens of memory sticks and disk drives in caddies all on USB hubs with wall warts? At the moment I don’t; I like a hard drive with backup. At the moment a whizzy SSD memory is too expensive

So can we please have a new fleet of proven design, known cost, nuclear stations and they will give sufficient time providing a working electrical grid to allow the energy storage people to progress – they will, and they will still get investment, because people are very keen on cars

14. Alex Terrell says:

For an upcoming report on 2050 demand and nuclear capacity options, I’ve also modelled wind and solar output compared to demand, and then put in an assumption for storage.

The result is that with 280GW of wind capacity, and 100GW of solar capacity, the UK would be running about 15% on gas, with a gas capacity of 95GW.

Vast amounts of storage can reduce the % of gas used, but make little impact on the required gas capacity. The reason is basically that when the vast amounts of storage are empty, you still need the gas.

This is the graph: http://prntscr.com/c7a7wc

I’ve also modelled demand over a day trying to respond to generation based on 14 March 2014, when wind started off at 6% and reached 61% later in the day, giving a healthy average of 36% over the day:

http://prntscr.com/c7ab0q
Even with extreme demand side response, that day would need 230GWh of storage. That’s just intra-day, before we get on to day to day storage.

People say: “storage costs are coming down fast, so we can use wind and solar”. I ask “how much storage do you think we’ll need”. Then I get no answer.

Storage at \$100/KWh is entirely feasible – and sound great. But does storage at \$100 billion / TWh sound so great?

So for a renewables infrastructure, it seems the optimal amount of storage would be somewhere around 100GWh, which would enable the gas plants to be run at optimum output, and avoid the issues faced in Ireland. But that still needs over 90GW of gas capacity.

• “The result is that with 280GW of wind capacity, and 100GW of solar capacity, the UK would be running about 15% on gas, with a gas capacity of 95GW.”

So you would need a gas grid the size of the current UK grid? Are you including electric cars etc?

• Alex Terrell says:

Yes, that is assuming heavy electrification of heating, transport and industry in 2050. Total demand on very cold days can reach 120GW.

• gweberbv says:

Alex Terrell,

did you take into account that UK is prioritizing offshore wind which has a higher CF? Probably this will not change the storage needs very much. But it might cut the installed capacity of wind turbines by a factor 2 or more.

• Alex Terrell says:

Hi, no I haven’t factored that in – that is a weakness in the model I should acknowledge (and I haven’t factored in other renewables). I don’t have data for offshore wind separated out from onshore.

I could extrapolate by years, but I few too few years – I didn’t want to go back earlier than 2012 due to low data sets. ,

Counter points could be that any future onshore wind will have a lower capacity factor as the best sites are gone. For offshore that isn’t the case – but they’ll get more expensive as they get further out to sea.

It won’t cut the capacity factor by two as the UK is already up to 27.5%:
Year Wind CF Solar CF
2012 27.8% 8.3%
2013 30.3% 12.4%
2014 26.1% 11.7%
2015 25.9% 11.3%
Total 27.5% 10.9%

• gweberbv says:

Alex Terrell,

here are the production data for offshore and onshore wind in the German grid for Jauary 2016 (blue=offshore, green=onshore).
What is important is the fact that offshore production has an on/off pattern. You either harvest near nameplate capacity or nearly nothing. In contrary, onshore wind produces most of the times around 30% to 40% of maximum production.

The result is that having offshore wind capacity significantly above demand will result in a lot of curtailment and very small additional production that is useful. This is a difference to onshore wind that goes beyond the simpel fact of a higher CF over the year.

I think it would be important to model this feature, if offshore wind is expexted to be the dominant from of wind power.

• robertok06 says:

“In contrary, onshore wind produces most of the times around 30% to 40% of maximum production.”

Sorry Guenter… but the graphs you’ve linked here DO NOT show anything close to what you state here.
“most of the times around 30% to 40%” means, to me, a wiggly line inside these two values… with few drops and highs… the plot you’ve provided is completelety different.

• Alex says:

Can’t check the details now, but I’ve heard German onshore wind CF is about 16%. That will get worse as wind turbines march southwards.

• donoughshanahan says:

@ GW, Roberto

I took wind data from pfbach’s site which reports MWh every hour for 2015, and came up with the following histogram (left column MWh, right column number of hours below MWh, so 76 hours below 500MWh)

MWh h
0 0
500 76
1000 286
2000 774
5000 2312
10000 2410
20000 2009
More 893

15. RegGuheert says:

What new battery storage technology at below US\$0.20/kWh REALLY does (when combined with current PV technology) is put a cap on how much electricity utilities can charge residential customers for their product. This is only true today at the lower latitudes (such as Hawaii), but in those locations it puts a cap of about US\$0.30/kWh on residential service. As utilities try to go significantly above this price, customers will migrate off of their network (or at least remove their loads and sources).

Another thing that battery storage does is limit that difference between peak and minimum TOU charges to about US\$0.25/kWh.

16. duffer70 says:

Why calculate storage capacity equivalent to a *year’s* worth of generator x? Isn’t this excessive?

17. A C Osborn says:

No on has mentioned Efficiency here, converting, charging, discharging and re-converting all have efficiency losses.
You can’t do it for nothing, so what will be the losses to the already difficult to achieve intermittent generation?
And why would you want to “waste” that electricity?
Nuclear, Coal & Gas every time.

• RegGuheert says:

Round-trip efficiency of Li-ion batteries is about 98%. One-way efficiency of modern three-phase DC-AC inverters is over 98%. Total round-trip efficiency for the battery and electronics is therefore over 90%.

For utility-scale storage, there are additional requirements for cooling pumps, etc., which brings the efficiency to below 90% but still above 80%.

• A C Osborn says:

I doubt Li-ion batteries will do it for Grid scale Wind Generation, but if it can then you are talking about a further 20% reduction in actual output.
When you add that to the required subsidies that Wind needs it really is an expensive way to produce electricity.

• RegGuheert says:

FWIW, even though I’m a fan of photovoltaics I’m simply not a fan of the current generation of wind generators, for a variety of reasons. Perhaps something like this is workable, but it needs more time to see how things work out:

http://www.kitegen.com/en/

• @RegGuheert,

but it needs more time to see how things work out

the time required may vary from one year up to never, not for technical or feasibility issues which are straightforward.
But because of a systematic sabotage of which I tried currently displaying and sharing the underlying mechanisms, here in the comments:

A job like KiteGen requires a minimum of approval and support as none has the economic resources, technical and above all psychological, to hold up racial hatred among the sources of energy.
The subsidies culture have generated millions of people looking to kill any truly promising energy innovations. The clearly promising innovations generate hundreds of people envious or ready to snatch the work and the ideas of others.

The attempt of denunciation of this annihilation of the commons, as is innovation, generates derision and underestimation of the complainant.

Now we have added a further problem, the last, and extremely convincing version of the full scale project including its detailed specifications and sub-assemblies validations, we will show only to those who understand, endorse, cooperate and sign an NDA, further provoking a public and persistent attitude of challenge and provocation against us.

It really used to drive me nuts having spin-offs, organizations and universities set themselves up as competitors to my company using the achievements we worked-out, done and shown, additionally using the taxes I paid.

Sorry for this rant, but Euan as you can see, miss to protect us from such slander attitudes coming from everywhere.

• Alex says:

98% round trip efficiency is an extraordinary claim and, as such, needs some evidence.

(a bit biased towards batteries?) gives 85%.
https://energymag.net/round-trip-efficiency/ gives 75% to 90%.
More info here: https://www.quora.com/Which-batteries-have-the-highest-round-trip-efficiencies suggesting below 90% and falling fast under non ideal conditions.
This http://www.cleanenergyreviews.info/blog/2015/11/19/complete-battery-storage-comparison-and-review claims 92.5% for the Powerwall, but that is DC to DC and under ideal charge, discharge conditions.

I think it’s prudent to assume, for pumped storage and Li-ion batteries, 90% charging, 90% discharging efficiency.

• Beamspot says:

Nope, real Li-pol batteries are in the 95% round trip ball park IF they work below 1C (low power, more than one hour to charge and another to discharge).

Depending on the type (hi energy like Tesla’s, then you have to be well below 1C, hi power, lower specific energy can do that at 1C or even 2C discharge/1C charge).

For home use, that means you usually need low power, so the 95 – 97 % roundtrip assuption holds. But not for cars where you need high power.

18. robertok06 says:

Interesting:

“Special Section on the Renewable and Nuclear Electricity: Opportunities, Challenges and Policy Recommendations”

19. Thinkstoomuch says:

RegGuheert,

“What new battery storage technology at below US\$0.20/kWh REALLY does (when combined with current PV technology) is put a cap on how much electricity utilities can charge residential customers for their product. …”

After all the talk about HI and capacity factor I thought I would look up current systems in the SMA public sites. Solar Edge only shows one. 🙁

I figure they are based on what the average folk would achieve.

In 2015 there were 33 sites that generated power. Average CF 16.51%

Well that can’t be right everybody is talking about 22% and that is 3/4 of that.

So then I went back to all the available data 2015-2010.

There were 124 outputs that averaged 16.36%. over that period. (similar to what Roger Andrews had done in the past, thanks for the thought fodder Roger. 🙂 )

Interesting information.

Not to mention that I would not recommend that the average folk doing an self install poking holes in the roof or things that use 100’s of amps. General 3k string would be 100 amps at ~30 volts the output of of a crystal solar panel)

So I really think you are being very optimistic on several fronts. It would be nice but …

I could fudge my way to build my own house and garage and all the things that go with it(currently helping build a half cement house complete with hauling the portland and gravel up 2000 feet from town). But I know I am in no way close to average or normal in my specific abilities (which are rather limited). But I am VERY leery of doing the above. Of course the last time I looked if doing a self install in FL I was looking at 8 cents plus just for the solar portion which is more than I pay per kWH. So not even a starter economically. Much less worth my time and effort.

What does a solar panel cost in HI anyway?

For what little it is worth,
T2M

• RegGuheert says:

@Thinkstoomuch (great moniker!)

“After all the talk about HI and capacity factor I thought I would look up current systems in the SMA public sites. Solar Edge only shows one.

I figure they are based on what the average folk would achieve.

In 2015 there were 33 sites that generated power.”

Interesting. Enphase currently shows 42,336 installations in the Hawaiian Islands:

https://enlighten.enphaseenergy.com/public_systems

“Average CF 16.51%

Well that can’t be right everybody is talking about 22% and that is 3/4 of that.

So then I went back to all the available data 2015-2010.

There were 124 outputs that averaged 16.36%. over that period. (similar to what Roger Andrews had done in the past, thanks for the thought fodder Roger. )”

Just for reference, I live in VA at 39 degrees North latitude and my 12.75 kWh Enphase-based PV array produces 18 MWh/year compared with 113 MWh/year nameplate for a capacity factor of 16%.

But my system is almost optimal: The roof is almost ideally pointed and there is no shadowing from trees or roof parts. In general, the roof on a house will not be pointed in the ideal direction, so the CF will be lower than that. In addition most roofs will be shadowed during significant portions of the day. Finally, string inverters from SMA are MUCH more sensitive to shadowing than microinverter systems. (SolarEdge has addressed this issue by putting “optimizers” at each PV module.)

Simply put, it’s not too surprising that the average capacity factor would be lower than the optimal value. With all the trees on Hawaii, it’s also not surprising the Enphase owns that market.

“Not to mention that I would not recommend that the average folk doing an self install poking holes in the roof or things that use 100’s of amps. General 3k string would be 100 amps at ~30 volts the output of of a crystal solar panel)”

Virtually no grid-tied PV systems involve “100’s of amps”. However, SMA systems often have DC voltages up to about 500VDC, which can be very dangerous.

Since I have microinverters, the highest DC voltage I deal with in my system is about 37V and the highest DC current is about 9A. Most of the system is 240VAC with currents up to about 45A (on a trunk cable between my main panel and the PV subpanel). The AC wiring is all very standard and can be done by any qualified electrician and by many homeowners.

“What does a solar panel cost in HI anyway?”

Frankly, with net metering, it makes very little difference what people pay for a PV installation. The reason is that residential customers in Hawaii pay between US\$0.35/kWh and US\$047/kWh for electricity:

https://www.hawaiianelectric.com/my-account/rates-and-regulations/average-electric-rates

Given that if a DIYer with a well-situated roof does it for US\$0.05/kWh or less then their system will fully pay for itself in about 5 years and then will produce electricity for the remaining 20 fully warranteed years for free. With the current federal tax credits, that payback time is about 3 years.

Someone who pays 3X that amount to have a system installed will have to wait 10 years to get their money back.

But with over 40,000 systems already operating on the islands, net metering’s days are numbered. Something’s gotta give. That’s where PF correction and battery systems come into play. Enphase offered PF correction (and customer-side monitoring information) as a way to shoehorn marginally more systems onto their grid. But that only gets you so far. As more and more people stop paying for the system, the price for those remaining will continue to rise. Eventually, customers will pay MUCH more to take energy off the grid than it costs to put energy on. (BTW, this is already the situation in parts of Australia. Consequently, that is where the first Enphase AC Battery and Tesla Firewall systems have been installed.

The good news is that first-generation battery systems are already below about US\$0.20/kWh. But that is per kWh STORED, not for every kWh used. In Hawaii, many of the loads can be made to follow the sun, so storage should be about 50% of consumption for a typical Hawaii home, meaning you would be paying about US\$0.10/kWh per ALL kWhs used by the home for storage.

• RegGuheert says:

BTW, this link indicates NREL’s current capacity factor for Hawaii is 0.17, not 0.22:

http://breakingenergy.com/2014/12/16/where-does-solar-make-sense-new-state-study-shows-big-differences-and-surprises/

That’s pretty close to what Thinkstoomuch came up with.

20. Olav says:

Hydro reservoars are also a battery and nature does the charging in traditional setup for free.
The next step is PHS which has the best performance of 1 KWh purchased electricity in and 0.75 KWh out when needed,. about the same as batteries has,
One very little used method is to pump additionally water from already high elevation into existing reservoirs. When doing so at rainy days is the enviriomental effect positive (less flooding). In Scotland I assume is rain and surplus wind going hand in hand. Many ponds could also be constructed for such a use and as 1 KWh purchased electricity goes in maybe 2 KWh can be delivered when needed if the additionally water can be found at high enough elevation.
Off cause is the scaling a problem as maybe 5% extra hydro in Scotland is tiny but it is a possibility which should be checked . 100 000 KWh of storage requires 1000 ton battery. The needed pond at 500m elevation does not have to be large and lasting 5x+ longer at a much smaller price.
To tide over rain lulls of a month or so is this kind of storage only possible for a small group of users. For the whole country is in absent of new nuclear fossil fuels needed but hydro reservoirs can be used to allow a more timely startup of gas and eventually coal plants.
Interconnectors to far away hydro reservoirs is a possibility but as I see from the Nord Ned Interconnector, it runs mostly flat out from Norway to Netherland and then it will not help at all in Netherland if some supply problem arises there. Interconnectors should the be used as a both direction backup with very little flow and that is bad economic for cable owner.

• “Tiny” Olav? Scotland’s mountains may be smaller, our rainfall may be less than Norway’s but science knows of energy storage methods for quantities of energy that are as big as Scottish ambitions in renewable energy.

Pumped-storage hydroelectric schemes are the next step I agree.

My proposal for a new pumped-storage hydro scheme in the Scottish Highlands, at Strathdearn near Inverness, could have an energy storage capacity of up to 6,800GWh.

World’s biggest-ever pumped-storage hydro-scheme, for Scotland?
https://scottishscientist.wordpress.com/2015/04/15/worlds-biggest-ever-pumped-storage-hydro-scheme-for-scotland/

Admittedly, 6,800GWh is not large compared to Norway’s hydroelectric energy storage capacity but it is big enough to serve Scotland and Britain well as the first (efficient but expensive) tier of a 2-tier energy storage system that can store all the surpluses from wind generated in Britain, on land and offshore, as my computer modelling shows.

Modelling of wind and pumped-storage power
https://scottishscientist.wordpress.com/2015/04/03/scientific-computer-modelling-of-wind-pumped-storage-hydro/

The second (inexpensive but inefficient) tier energy storage of a 2-tier system will use power-to-gas to store almost unlimited amounts of energy.

Power-to-gas can make hydrogen, methane or syngas, which can be turned into synfuels.
https://en.wikipedia.org/wiki/Power_to_gas
https://en.wikipedia.org/wiki/Synthetic_fuel

Storage of hydrogen and methane underground is well established and synthetic fuels are stored just as easily as petroleum-based fuels.

All that is plenty – more than enough energy storage capacity – but I think science can offer one more new option too – my proposal for Deep Sea Hydrogen Storage research and development of a simple offshore power-to-gas energy storage method.

Off-Shore Electricity from Wind, Solar and Hydrogen Power
https://scottishscientist.wordpress.com/2015/04/23/off-shore-electricity-from-wind-solar-and-hydrogen-power/

Science has all the answers for energy storage but what really needs to proceed apace now is new political leadership to overturn the negligent failure of successive UK governments who have refused to borrow to invest in building the necessary new energy infrastructure.

With regard to the Telegraph article linked to – “Holy Grail of energy policy in sight as battery technology smashes the old order” –

We don’t really need any breakthroughs in battery technology, (interesting and useful those may be).

Norway has been supping from the “Holy Grail” of hydroelectric power and so Norway’s economy benefits from “eternal life” – environmental sustainability.

So actually, the old order technology of pumped-storage hydro and power-to-gas work perfectly well to store all the energy we will ever need to store.

Rather it is the old order of fiscal conservative UK governments, who hate to borrow to invest for the country’s and the people’s needs, that really needs smashing.

Scottish Scientist
Independent Scientific Adviser for Scotland
https://scottishscientist.wordpress.com/

• Alex says:

Your proposal is certainly ambitious (and also un-costed and probably unacceptable on environmental grounds), but if we look at 2050 demand, then 6800GWh is less than 4 days winter demand.

It would reduce the gas utilisation by a lot but make very limited impact on the gas capacity required.

• Accurate cost estimates come later in the life of a project, after a properly resourced design team is commissioned. For now, at concept proposal phase, only inaccurate costs estimates can mentioned in passing, so I offer the following comments on costs.

Link to google – cost of pumped-storage hydro

In 2013, the BBC reported that the SSE had proposed a 600MW / 30,000MWh pumped-storage scheme for Coire Glas, Lochaber at the estimated cost of £800m.
http://www.bbc.co.uk/news/uk-scotland-highlands-islands-25365786

From the SSE’s estimate I extracted 2 approximate cost factors for pumped-storage hydro schemes in Scotland …
£1.33 per Watt (W)
£0.0267 per Watt-hour (Wh)

which factors I used to make a first order estimate in another blog post of mine

“Scotland Electricity Generation – my plan for 2020”
https://scottishscientist.wordpress.com/2015/03/08/scotland-electricity-generation-my-plan-for-2020/

“Total pumped-storage power – 6GW, energy stored 160GWh, – cost £4.3 to £8 billion” (the cost of the pumped-storage scheme alone, not including the higher cost of wind turbines)

Why 160GWh? Well from this (again)

Modelling of wind and pumped-storage power
https://scottishscientist.wordpress.com/2015/04/03/scientific-computer-modelling-of-wind-pumped-storage-hydro/

“Such modelling can predict how much wind power and pumped-storage energy capacity should be installed for satisfactory renewables-only generation.”

“I used estimated relative costs of wind turbines versus pumped-storage hydro to arrive at a optimal minimum cost balance –

store energy = 1.11 days x peak demand power
annual maximum wind power = 5.5 x peak demand power”
_______
For peak demand power of 6GW

Store energy = 1.11 days x 6GW = 6.66 GW-days = 6.66 x 24 = 160GWh

Annual maximum wind power = 5.5 x 6GW = 33GW
(costing another £45 billion for the 28GW more on top of the 5GW Scotland already has, that’s just for Scotland!)

Looking at the very high cost of all the wind power required, the cost of peak demand back up from gas CCGT is relatively modest.

It is probably a good plan to build the gas capacity back up first, before building all the pumped-storage and wind turbines required, because at least we could be sure that the country can afford the gas capacity required.

Which is going to look like a dash for gas and have some wondering if there really is a plan to go for 100% renewable energy but there is indeed, because those gas CCGT power plants can burn hydrogen or methane or syngas from power to gas as easily as they can burn natural gas.

My inaccurate cost estimates indicate about 10 times as much needs to be spent on wind turbines as on pumped-storage hydro, at today’s prices.

Of course, that’s modelling a simple 1-tier energy store.

Using a 2-tier energy store, backing up with power-to-gas, I expect would lower overall system costs but my how much I can’t say because I haven’t had the time to do that much more complicated modelling task.
______
1.11 days of peak demand power is about 1.6 x 1.11 or 1.77 or less than 2 days of average demand power.

Energy storage of 4, 3, or even 2 “days of winter demand” is not needed, not if enough annual maximum wind power is installed – 5.5 x peak demand.

Using those same cost factors, I can estimate, here and now, the back-of-a-fag-packet estimated costs for an energy store for 100% renewable energy operation in Britain

Some more examples, with rounded arithmetic
______
* for a 50GW / 1,400GWh scheme for Britain 2020 that would cost approximately
£37 to £66 billion (the cost for pumped-storage alone, wind generators maybe x10 that)
_____
For peak demand power of 50 GW

Store energy = 1.11 days x 50GW = 55.5 GW-days = 55.5 x 24 = 1332GWh

Annual maximum wind power = 5.5 x 50GW = 275GW
_______

For peak demand power of 52.5 GW

Store energy = 1.11 days x 52.5GW = 58.3 GW-days = 58.3 x 24 = 1400GWh

Annual maximum wind power = 5.5 x 52.5GW = 290GW
_______
* for a 120 GW / 3,360GWh scheme for Britain 2050 (when all heat and transport is electrified) that would cost approximately
£88 to £158 billion. (the cost for pumped-storage alone, wind generators maybe x10 that)
____
For peak demand power of 120 GW

Store energy = 1.11 days x 120GW = 133.3 GW-days = 133.2 x 24 = 3200GWh

Annual maximum wind power = 5.5 x 120GW = 660GW
____
For peak demand power of 126 GW

Store energy = 1.11 days x 126GW = 140 GW-days = 140 x 24 = 3360GWh

Annual maximum wind power = 5.5 x 126GW = 700GW
___

So if – and admittedly it is a big IF – if it is possible for Britain to install 700GW of annual maximum wind power then 3360GWh would indeed be enough energy storage capacity, which is well within the scope of the Strathdearn pumped storage hydro scheme proposal.

But you know, that’s 2050 and a lot can happen between now and then. 700GW of wind power – wind generators spread across the Atlantic – may be a doddle by 2050?
__

Costs vary a lot because pumped-storage hydro schemes vary a lot, in power, in energy stored, in their geographical setting which determines unique design elements.

Unique variables that will vary the costs for Strathdearn are the dimensions of the dam and the canal, which depend on the design power and energy storage capacity.

Basing my inaccurate cost estimates on Coire Glas prices is likely to be very inaccurate, especially for any scheme like Strathdearn which is very unlike Coire Glas is its unique features.

The British government could borrow, courtesy of the nation’s savers, the £10s of, or £100+ billion the pumped-storage hydro Britain needs might cost, investing in it as national energy infrastructure, while costing taxpayers and electricity customers nothing.

Remember Mark Carney, the Governor of the Bank of England has been increasing the money supply with new money of £100s of billions created in Quantitative Easing (QE), with not much to show for it, apart from inflated asset prices.

I propose instead to borrow similar sums of money to pay for energy storage. Carney’s QE cost taxpayers and electricity customers nothing. Neither would government borrowing to invest in energy storage for the national grid.

Of course, there MAY be a problem with fiscal conservative Tory Chancellors who just love to reduce the deficit, hate to borrow, refuse to invest in infrastructure and threaten to deflate the UK economy into a slump, requiring Bank of England QE to bail a fiscal conservative chancellor out of the hole he has been digging the economy into.

For years, the infrastructure investment problem was Cameron – Osborne.

Now the problem would appear to be May – Hammond, unless they are intending on surprising us all soon by announcing £100s of billions of infrastructure spending?

• Alex says:

The three gorges dam is 2335m long and 120m high, and cost \$27 billion in 2003. Your scheme is similar length, but 2.5 times higher, so 6.25 times the volume. Given construction costs might be double what they are in China, and add in inflation, we could estimate around \$500 billion.

Then we need the HVDC cable costs – say 80GW over 500km – perhaps £40 billion. Plus the canal, perhaps a total of \$600 billion for the 6.8TWh scheme.

If the holy grail of battery storage is \$100 billion / TWh, then your scheme might be cheaper then batteries.

Alternatively, \$500 billion will buy at least 85GW of nuclear capacity meaning we wouldn’t have to bother with storage.

• There’s no sugar-coating the fact that the Strathdearn pumped-storage hydro scheme proposal sacrifices –

* the upper glen of the River Findhorn from source as far downstream as Invereen, the river would be obliterated, with downstream from Invereen severely reduced river flow implications,
* Loch Moy and the route from Loch Moy to the Firth of Inverness, or alternatively to further up the coast, to the Firth of Moray

– all for the greater environmental good of the country and of the world – helping to bring an end to fossil fuel burning power stations throughout Britain, whose carbon dioxide pollution would otherwise cause even wider and more unacceptable ecological problems.

The motivation for renewable energy schemes is an ecological motivation. The world is transitioning to green energy in order to save the planet from global warming, as was agreed at the UN Paris Climate change conference recently.

It is the great cause of our age and some must be asked to make sacrifices bravely.

If the upper River Findhorn, Loch Moy etc are sacrificed for the greater environmental good, it means that many other rivers and glens in Scotland can be spared the destructive hydro scheme works that would otherwise have to be considered.

Whilst I would expect a certain amount of NIMBY resistance to the Strathdearn pumped-storage hydro scheme, the decision about whether it goes ahead or not would be a decision for government, taking into account the wider public interest as they do with other planning consent applications.

• The superficial structural volume of the Strathdearn Dam, super-sized to the full 6,800GWh would be about 80 million cubic metres

https://scottishscientist.files.wordpress.com/2015/04/strathdearndam.jpg

which is only 2.9 times not “6.25 times the volume” of the Three Gorges Dam at 27.4 million cubic metres and 181 metres high, not “120m high”
https://en.wikipedia.org/wiki/List_of_largest_dams_in_the_world

so I suggest that you halve the number you first thought of, to arrive at your estimate for the super-sized 6,800GWh dam that is twice what Britain needs in 2050 [\$250bn? / £190bn?]

then maybe halve it again to arrive at your estimate for the 3360GWh dam that is what Britain needs in 2050 [\$125bn? / £96bn?]

then maybe divide by 2.4 to arrive your estimate for the 1,400GWh dam Britain’s needs for today’s 50GW+ peak demand [\$52bn? £40bn?]

– assuming as a first approximation that dam structure volume is proportional to energy stored, which it won’t be exactly.

The recent scheme to double the capacity of the Suez Canal cost only \$8.4 billion
https://en.wikipedia.org/wiki/Suez_Canal

which required 260 million cubic metres of dredging and 250 million cubic metres of dry excavation
http://www.suezcanal.gov.eg/sc.aspx?show=69

which sounds like more work than the Strathdearn Power Canal volume of 216 million cubic metres, for the full sized 255GW but maybe similar because there would be a lot of ground levelling that needs to be done to allow for one continuous canal the full 30km length.

So say the same as Suez 2, \$8.4 bn for the 255GW power canal.
\$4 billion for the 120GW Britain 2050
\$1.6 billion for 50GW Britain 2020

The power canal may only be 3% of the cost of the dam.

With respect, I don’t think your costings are so accurately done that they can disprove my initial inaccurate estimates based on Coire Glas prices, which to recap were

* for a 50GW / 1,400GWh scheme for Britain 2020 that would cost approximately
£37 to £66 billion

* for a 120 GW / 3,360GWh scheme for Britain 2050 (when all heat and transport is electrified) that would cost approximately
£88 to £158 billion

But we can bandy figures back and forth without ever arriving at an accurate estimate because there is not even the details of a full design specified with which to make cost estimates.

Anyway, I think we have filled up the back of an envelope with numbers, so that’s as much as can be expected at this stage.

Thanks for your interest Alex.

21. Leo Smith says:

I find it very depressing when a hitherto good site discussing things rationally gets taken over by adverts and shills from the very thing it exists to critique.

after years of sunbsidised development we know that windmills cost at least twice as much as nukes, and solar even more, and with storage neither will cost less.

the renewable shill starts from the assumption that his target market should consider that nuclear is unthinkable and fossil is obscene: ergo fluffy renewables must be made to work irrespective of cost.

Challenge that primary assumption, and the whole renewable industry is revealed as pure fraud.

• gweberbv says:

Leo,

what do you think about the recent auction in Chile?

12 TWh per year, for 20 years, starting from 2021, \$47.59 per megawatt hour (average price)
Lowest bid: The Spanish solar-energy developer Solarpack Corp. Tecnologica won a contract to sell power from a 120 megawatt-solar plant for \$29.1 per megawatt-hour.

(Note: Has nothing to do with batteries.)

• Leo Smith says:

I would say that someone who auctions for power without specifying when it is to be delivered is insane or has been bribed.

• gweberbv says:

Of course the delievery time (‘block’) is specified. It is up to the bidders to make sure they can deliever. If their own plants are not delievering, they have to buy electricity on the spot market or they have ramp up backup plants to fullfil their contracts.

But if you have PV plant sitting in a desert with very few clouded days, it should be not terrible complicated to deliever during the 11 a. m. to 3 p. m. block (just as an example).

As far as I know Chile has such auctions since about a decade. So, there is plenty experience how to do it.

• robertok06 says:

Chile has experience… but also has lots of problems to solve, before PV can become a real large scale source of electricity:

“Chile has four main separate networks, which means that the high penetration of utility-scale solar in the desert regions of the north cannot be used to benefit the most populated central regions around Santiago until the northern and central grids are interlinked. As a result major project is underway to connect the Northern Interconnected System (SING) grid with the Central Interconnected System (SIC) grid. This is due for completion in mid-2017.”

… this was published in April of this year… who do you think should pay for the large costs, and construction times, of interconnectors and long-distance electricity transport lines?

• robertok06 says:

“12 TWh per year, for 20 years, starting from 2021, \$47.59 per megawatt hour (average price)”

It’s a fraud.
They bid low now hoping that the cost decrease will continue unabated forever… SunEdison has gabled the same way and SunEdison has gone bankrupt, in the very profitable market of Hawaii.

I really can’t understand how a person normally endowed like you, guenter, cannot get the simple fact that 47.59 dollars per MWh ONLY WHEN THE SUN SHINES, and leaving ON OTHERS, at THEIR expense, the technically/economically challenging task of balancing power when there is too much sunshine (and/or low demand) or there’s no sunshine at all (more than 50% of the time by default).

Why? What’s so difficut to understand, guenter? Tell me.. I can help.

R.

• gweberbv says:

Roberto,

the 12 TWh are a good portion of the annual consumption of Chile. It does not only affect the time during the PV peak. When you look from where Chile is coming in terms of electricity prices, they will be very happy to build a few transmission lines if this allows to cut their power prices by a factor 2.
If this is fraud, it will be very expensive for Italien tax payers. Because Enel won about 40% of that 12 TWh (starting from 2021).

• robertok06 says:

“If this is fraud, it will be very expensive for Italien tax payers. Because Enel won about 40% of that 12 TWh (starting from 2021).”

Guenter: the extra cost eventually borne by this fake bid is nothing compared to the 6,7 billion Euros/year for 20 years that the italians are already paying for a mere 23 TWh/y… around 29cEuro/kWh until 2033 (but “no nuclear, thanks… because it is very expensive… 92 pounds/MWh”… how funny is that?… the italians are used to it, don’t worry. 🙂

• JerryC says:

“It is difficult to get a man to understand something, when his salary depends upon his not understanding it!”

– Upton Sinclair

• Shills? Your evidence for that? I would say you were the shill.

22. clivebest says:

There is a very simple solution to supply reliable renewable energy in flat countries like the Netherlands or the UK. Simply build ~100km wide upturned umbrellas to collect rain water for hydro power.

23. Apollo says:

When Hinkley Point is called dispatchable energy it becomes kind of obvious the doomer discourse has cornered itself.

Indeed there has been a visible decline in the costs per kWh of Li-ion batteries. Anyone trading with goods such as cellphones, portable computers or electrical vehicles has experienced in the past 4/5 years an increase of storage capacity by a third and a decline of weight by a fourth for the same cost. However, before the end of this decade, Lithium itself will become more than 2/3 of the costs of these batteries and this lovely trend will come to an end.

The optimistic discourse is banking on different technologies, with metal-air and flow batteries at the head. Some of these are well advanced in their lab-to-market process; it is not hard to envision batteries under 200 \$/kWh for sale before 2020. None of these technologies are bound to make a huge impact on transport (Li-ion will still be better on weight and density) the challange they pose is to the centralized grid. In the US this problem is known as grid defection, and if Tesla was able to do what it did with such a lousy product, imagine the impact of a battery going for 150 \$/kWh.

People in Europe should take a closer at what happened in Spain the past couple of years. The government went to unimaginable lengths to prevent grid defection. And there is no guarantee the various judicial process that resulted from it wont leave the centralized grid unprotected at some point.

• Beamspot says:

I have many spanish colleagues that live totally off grid.

One of them even calculate and deploy such systems.

Based on their experience, there are no legal issues to become off grid, grid defection, it was really easy for them to go down that route.

The problem they experience in real life, is that electricity is REALLY expensive, so they tend to use as little as they can, they moved everything they could out of electricity, excetp air conditioner by the simple reason that PV gives enough energy in summer time just at the moment that the AC is needed.

Wintertime usually work by diesel generators, with the excess heat used as central heating (much higher efficiency), and they even have Solar Hot Water (the claim, all of them, the best investment they did, by very far).

All people I know that live with PV/wind and off grid, are really unhappy and even against renewables.

Only those that sell their PV electricity to the grid well above parity are happy, and those are the kind that are so angry because selling it at par didn’t give them any profit.

BTW, those claiming selfconsumption by PV/grid attachment tend to be in the high class, not in the middle or lower class (that don’t have possibility to finance something that can’t be placed in the little space they have available and that generates almost nothing). This is another scam hiding wealth pumping from lower classes to higher classes.

• ristvan says:

All forms of metal air have yet to solve the dendrite formation reliability problem. VERY difficult since inherent in the metal side electrochemistry, no matter the transition metal.All flow battery chemistries have yet to solve cost and cyclelife issues. The latter seem inherent in corrosive chemistries, purity issues, and such. One cannot rule out a miracle, but that is what it will take. See new comment downthread from this semiexpert.

• robertok06 says:

The last flow batery I’ve seen, from a German company, weights 2.5 tons to deliver the electricity needed by a Nissan Leaf to run 500 km…

24. ristvan says:

The only present commercial grid storage battery is sodium sulfur molten salt running at 450C. NaS are very expensive, and primarily used for peaking on thinly transmission supported remote small towns where local gensets or a second or beefed up transmission line would be still more expensive. Presidio Texas (installed 2010) was 4MW x 8 hrs at a cost of \$25 million; lifetime ~15 years. Largest single grid storage battery in the US at present.
PBA hasn’t grid cyclelife; the one company (Xtreme Power)which tried to achieve it via a variety of tricks went bankrupt after installing 60 MWh. LIIon is too costly, despite Musk’s Gigafactory bet. A123 went bankrupt after installing one 20MWh unit (for frequency control, not bulk storage) using a \$17.1 million US loan guarantee. None of the flow chemistries (conceptually good for grid) have achieved either cost or cyclelife, although this still an area of active research. Idea of converting e to fuel and back to e (usually via some hydrogen scheme involving some form of fuel cell) is too thermodynamically inefficient to be practical. Supercaps have sufficient cyclelife and power density, but one order of magnitude too litte energy density even after my materials ~ 50% improvement (which also cuts their cost ~25%). Beacon high speed flywheels were too expensive, unreliable, and insufficiently energy dense. Beacon went bust after one frequency control installation–again not intermittency bulk storage. Having researched electrochemistry for now 11 years, resulting in several issued patents concerning improved supercaps, my opinion is that after pumped hydro, grid storage is a chimera except in special cases for NaS. And that means renewable intermittency is economically fatal at any significant penetration, because inefficiently utilized FF backup investment is required (unless one has Norwegian hydro to hand).
Essay California Dreaming in ebook Blowing Smoke covers the grid storage waterfront as of ye2014. Essay Hydrogen Hype discusses all the hydrogen scheme issues.

25. PhilH says:

I’d be interested to see a line on the figure for the storage required for a nuclear scenario:

GB’s demand, very roughly, varies between 20GW on summer weekend nights and nearly 60 GW on winter weekday evenings, say an average of 40GW. To have nuclear stations supply the 21st GW and more means they are used less and less up to the 60th GW, which is wasteful and increasingly expensive, perhaps ruinously so.

How much battery storage would be required to even out the demand over the year, using 40 GW of 24h/365d stations, instead? A trivial initial estimate would be 80,000 GWh, but a more sophisticated analysis, taking account of the daily highs & lows of the spring & autumn, ought to give a figure substantially less than this – but by how much? I’d be very pleasantly surprised if it were less than 8,000 GWh, but I’d like to be proved wrong.

• Alex Terrell says:

I’m looking into that (with Andy Dawson) for 2050, rather than now. We think we can do it with 3% gas and no significant storage.

Report to follow.

• PhilH says:

I was assuming the lines in the figure were for no FF and all storage, so that’s what I’m interested in for nuclear for comparison with the all-solar and all-wind scenarios. Presumably, the more FF you add in to the solar & wind scenarios the less storage you need, as well.

• Alex Terrell says:

The more storage, the less FF usage, but there is not much impact on FF capacity.
http://prntscr.com/c81bv8 (this is based on 2050).
When the storage is empty, it’s empty and you need close to full capacity backup. The consequences of storage being empty and having insufficient backup could make Fukushima look like a blown light bulb, so you’d have to cater for 1 in 1000 year weather patterns. (Or why no 1 in 10E6 years – which is the sort of requirement for reactor core damage frequency?)

We haven’t modelled storage for nuclear, but are looking at 613TWh of nuclear per year and 20TWh of gas (or diesel or bio-diesel) per year. So 20TWh of storage is the upper bound, assuming all cold days are lumped together.
http://prntscr.com/c81dl4

This assumes that daily demand can be flattened, or that there is sufficient storage to do this. The main source of “daily storage” in winter is the thermal mass of buildings.

• clivebest says:

If we had 50GW of nuclear plant running 24/7 then there would be free power available at night. We will need this energy to charge all the electric cars overnight, and also to store heat/cold for release during the day in buildings. Currently electricity is only about 33% of UK primary energy needs. If we are serious about decarbonising transport and heating as well then we will need all the nuclear power we can possibly get. We probably will always need a reserve of ~15GW of Gas to cover downtimes and extreme demand in winter.

• robertok06 says:

JUst look at France!.. they have 53 reactors in 19 power stations, with 63 nominal GW powe… with a bit less than 7 GW power available as pumped-hydro stations.

I’ve not found the data about production from pumped-hydro (PH) for a full year… but assuming a maximum CF of 50% (irrealistic, it would mean that PH is 100% of the time either pumping up water of producing electricity.. 7 GW at 50% CF correnspond to 3.5*8760~ 30 TWh/year.
All this for 410 TWh/y from nuclear.

26. BMD says:

Robertok06 some statitsics for you !