By Rud Istvan
Got to thinking about WUWT 2.0, and some of the fact ammo climate skeptics here might need if/when Kerry becomes Biden’s climate guru. So here is a bit more WUWT fact ammo for the maybe coming renewed climate war.
One of the BIG problems with renewables is their intermittency. Another is their lack of grid inertia. (See my recent guest post on Grid Stability for details.) Innumerates like Kerry persistently claim both issues can be/will eventually be overcome by more grid interconnectivity or by better grid battery storage. Those hopes/beliefs are almost certainly wrong. My last post, referenced above, challenged the interconnectivity belief. This hopefully not too technical complementary post explains why their rechargeable grid battery hope is also wrong. It does so in a simplified yet easily researchable way. It provides all the key words for anyone seeking deeper grid battery storage understanding.
There are basically only three non-kinetic electrical storage mechanisms: electrochemical, pseudocapacitive, and capacitive. We deliberately exclude indirectly kinetic pumped hydro, for which there are two big expansion problems: insufficient remaining suitable terrain, and California having rejected it, period.
Electrochemical batteries involve some chemical species change (lead acid aka PbA is discussed following, chemically oversimplifying, the lead anode goes to lead sulfate then back to lead thanks to the sulfuric acid electrolyte). The general nature of such reactions is that they are faradic (involving electron exchanges, named in honor of Michael Faraday, also honored by the Farad of capacitive charge). The original faradic battery discovery was by Volta centuries ago (the volt, unit of DC electromotive force–EMF, is named in his honor). He used zinc, copper, a paper separator, and saltwater electrolyte to twitch frog’s legs with electricity from his ‘battery pile’. The same ‘electrochemical erosion’ principle is used today by sacrificial zinc anodes to prevent hull corrosion in saltwater.
You can build a modern ‘Volta Pile’ replica sufficient to light a single mini Christmas bulb (for a while) just using two US pennies and a lemon. Sand one of the pennies to remove its copper plating (copper is now so valuable, pennies are only copper plated zinc, yet the US treasury still loses money on each one minted). That gives you Volta’s zinc and copper electrodes. Insert both mostly into deep paring knifed slits about ¼ inch (max 1 cm) apart in an unpeeled lemon (pulp is separator, acidic juice is electrolyte). Attach two wire clips to the protruding penny edges and then to the single mini-bulb leads (completing a DC circuit where polarity does not matter). A low voltage DC current flows sufficient to light the resistively heated mini-bulb, until the copper plating on the unsanded penny is electrochemically eroded. Makes a great entry-level high school science fair experiment. This easy home experiment also illustrates one of the many ways all batteries, rechargeable or not (as here), have limited lifetimes much shorter than grid lifetimes.
Pseudocapacitance is rare and complicated. We mostly skip it. There were some DARPA funded ruthenium hydroxide based efforts a decade ago, all failed. Evans Capacitor has sort of one based on tantalum, but Dave’s expensive high power military device for single side band fighter aircraft radar power pulsing is actually an electrolytic cap (defined below) with pseudocapacitive ‘overtones’.
Capacitive storage comes in two flavors: all ‘solid state’, or across some ‘Helmholtz’ layer. All ‘solid state’ is most of electronics billions of little ‘chip caps’ today, just two metallic conductors separated by a ceramic dielectric. The Leyden jar was their progenitor. Even ‘wet’ aluminum electrolytic capacitors susceptible to electrolyte drying failure are technically still in this ‘solid state’ flavor. Very fast, almost infinite cycle life (except for wet dielectrics), but very faradic low charge storage. Dave Evans’ brilliant tantalum military stuff is still just a very high power density version of the wet form of ‘solid state’. All caps of this ‘flavor’ have grossly insufficient energy storage to complement grid renewables.
The other flavor uses the Helmholtz ‘electrolytic double layer’ capacitance effect. The most familiar example is thunderstorm lightning, which Helmholtz first explained (the ‘electrodes’ can be vapor to water, or water to ice—probably both in any big Tstorm cloud). The commercial example is a supercapacitor (aka EDLC). Cycle life is a couple of million full discharges. Used in power dense applications (think Navy rail guns), but at best only about 1/10 the energy density of a high power LiIon battery. They are used in grids as ‘statcoms’ up to about 4 MW, primarily for power factor correction and related grid frequency support. Not nearly enough EDLC energy density for renewable intermittency grid support.
Green woke Formula 1 racing tried, then dropped, both high power LiIon and EDLC hybrid designs. Too many car battery fires thanks to lack of sufficient LiIon power density, while EDLC were (prior to my issued carbon materials patents) physically too big to easily fit the car racing chassis at F1 horsepower. F1 finally dropped their woke green hybrid racing experiment totally—bad for business. Real world intruded. Below is a hybrid Mercedes F1 in Singapore, after a practice lap, after its high power LiIon blew up with the driver sitting on top of it.
As a potentially complicating (and hopeful future) side note, a speculative post over at Judith Curry’s Climate Etc some years ago concerned Henrik Fisker’s second electric car venture. There is a hybrid half battery/half EDLC device variant (one electrode of each type) melding attractive properties of both (high energy and power density, 20000 cycle life). It is in limited production for things like large inductive motor power factor correction. Invented then sold by Subaru since not good enough then for hybrid EVs. As yet not commercialized for EV’s save for Fisker, and even he deferred to LiIon for his first ‘Fisker2’ vehicles.
Battery limitations
All energy storage devices are characterized by three basic physical parameters ignoring cost: (1) energy density (how much charge they hold, aka how long they take to discharge, in Wh), more is better, (2) power density (how many charge/discharge Amps per second), higher is better, and (3) rechargeable battery cycle life, where higher is always better.
It is fairly easy to get 2 of 3 of these in any commercial rechargeable battery system. Getting all three in one device is REALLY hard.
For example, PbA (2+3) are car starter batteries. PbA (1+3) are golf cart batteries. ALAS, they are NOT interchangeable. The former uses thin electrodes for power. The latter uses thick electrodes for energy. The cycle life of the former is about 2x greater than the latter since the former (by definition) do NOT normally discharge nearly as deeply as the second. Deep discharge always chemically deforms the PbA electrodes more rapidly via (large crystals) sulfation and thus shortens cycle life. A starter battery is dead from sulfation after about four full discharges. Is also why older starter batteries usually die in winter.
The degree and rate of discharge/recharge has a profound impact on battery cycle life for several reasons. Simply put, ‘full’ charge/discharge EV batteries of the same chemistry will last much less than hybrid EV batteries like in my Ford Hybrid Escape, MY 2007, NiMH chemistry, now 13 years old and still going mostly strong (engine start exception after sitting a week due to slowly increasing leakage current, which Ford foresaw by providing a ‘jump start’ [not really] button to use as the battery ages and the car sits more), because the hybrid traction battery charge always floats only between about 45% and 55% of full charge.
BTW, my hybrid Escape made imminent economic ‘green’ sense. It is a small but sturdy I beam SUV, AWD with class one tow hitch, total HP ~210 (comparable to the 3 liter V6 variant). The HP comes from a downsized 1.4 liter Atkinson cycle I4 at 140 HP, and about 70 HP equivalent from the electric machine. (Atkinson cycle sacrifices torque for about 15% fuel efficiency over the Otto cycle—but it doesn’t matter because the electric machine more than makes up any torque deficit). The MY2007 V6 got about 22mpg highway and about 18mpg city. Our hybrid version still gets about 32 city and 28 highway (well, 27 at our usual 75 mph with summer AC on). The hybrid premium over the V6 was almost exactly $3000; in 2007 the hybrid federal income tax credit was about $3000 (not a coincidence, a Ford pricing strategy). So we were making money from fuel savings on the day we drove the car home from the dealer. Best part is, our I4 Atkinson uses regular gas; the equivalent Otto V6 needed premium. Where we are, the octane difference is over a dollar a gallon. Fewer yet cheaper gallons.
By far the most numerically common rechargeable battery now is Lithium Ion (LiIon) aka the lithium rocking chair, invented more than 30 years ago. It is in a sense partly pseudocapacitive via its ‘rocking chair’. Lithium ions migrate on charging from their chemical home metal cathode, thru a Lithium Ion electrolyte, to intercalate into the carbon anode (the rocking chair). Intercalation involves no chemical change to the anode, just lithium ions ‘snuggling up’ to their ‘rocking chair electrons’ inside the carbon anode. Only the metallic cathode (where the cobalt is) changes chemically with charging/discharging.
LiIon energy/power limitations are similar to PbA. Making everything thinner gives more surface area per battery volume for maximum energy density. A Tesla EV battery is big enough for range that power density is not a primary consideration except in charging. Its cycle life limitations are different but still constraining. F1 LiIon made things thick for power density—just not enough.
There is a secondary relatively low power density consequence for Tesla. Rapid charging generates more heat than can be easily conducted away. The Nernst equation says that heat kills cycle life (above ~40C, about 2x per 10C). Something Tesla does NOT say about its rapid charge stations. You can rapid charge often for convenience, but doing so will also kill your car’s Tesla battery quite early. Perhaps an undisclosed Tesla financial warranty liability?
Of course, as the ‘stunt’ Tesla grid battery installation in South Australia shows (more about it follows), it is NOT economic and cannot hold up the grid for very long. And since Tesla introduced its home grid ‘Powerwall’ s few years ago, the price has increased (not decreased, per Gigafactory promises) about 35% while the warranty was cut by about 1/3. Not a good economic deal.
Future battery possibilities
These tend to come in three hopium flavors.
First, nanotech will come to the rescue. Except all the examples to now were either frauds (Silurgy, NanoOne), or failures like A123 Systems, or speculations without even lab proof of concept like a recent WUWT post about lithium sulfur.
Second, flow batteries for the grid (where the charge is stored in the liquid electrolyte rather than in the electrodes, so with sufficiently big electrolyte tanks energy density is theoretically unlimited). Except all these various flow chemistries have commercially failed to date despite California encouragement and much VC investment. Low cycle life and/or high cost (bulk vanadium isn’t cheap, and cheap rhubarb hasn’t cycle life). See essay California Dreaming in ebook Blowing Smoke for several (now a bit dated) specific illustrated examples.
Third, exotic chemistries like sodium sulfur work, but are expensive and very high temperature. Lithium sulfur (a recent WUWT post on new theories about how to solve the two problems that still exist) are without even yet even lab proof of concept because of the inherent engineering difficulties of making one.
Two final definitional climate warrior ‘ammo’ reminders
First, batteries live in a DC world. Grids live in an AC world. There is always the significant added cost and limited reliability of the necessary high voltage high power DC/AC interfaces. Quoting battery cost without the unavoidable interface cost is intellectually dishonest. Tesla Powerwall 2 came in either DC only, or DC/AC integrated. The price difference in 2017 was about $1500 on a 14KWh base DC battery then about $5500. So about plus 30% for grid interface costs.
Second, using batteries to solve renewable grid intermittency is something touted by Elon Musk, and ‘gifted’ by him to South Australia after their 2016 renewable induced disastrous grid blackout. But Elon used a simple marketing ‘con’, same as the California flow batteries exposed in essay California Dreaming.
It is perfectly possible to truthfully specify a grid battery installation in MW. After all, it has those. BUT the grid intermittency relevant value is MWh (how long the battery lasts providing those vital MW of backup electricity). When the MWh answer is minutes while the grid MWh need is hours, the installed grid MW battery capacity joke is on you while your grid goes dark.
Elon ‘conned’ South Australia (IMHO to gain free ‘advertising’), and Australia’s MSM never caught on. Tesla’s Hornsdale, SA facility (below, now expanded by 50%) delivered 150 MW! But only 189MWh. It can hold up the SA grid for little more than an hour. The South Australia blackout duration depended on where you were; metropolitan Adelaide was restored first. Central Adelaide was dark for at least three hours. More symbolic hopium that doesn’t work in the real world.
Our municipality sells very affordable private/hobby logging permits. Namely the right, with a few simple rules, to clean up for firewood whatever professional loggers leave behind in designated easy to access forest areas.
My choices for a happy day in the woods are: -A gas powered chainsaw with 40 minutes autonomy between refills and a 5 liters jerrycan. Total cost, about 250 bucks for a nice powerful refurbished one with 3 years warranty.
Or a 400 bucks mid-range cordless chainsaw with battery autonomy of 7 minutes and half a dozen of spare batteries costing 170 bucks each. Even the option of a gas generator to recharge on-site.
Each and everyway you do the maths, the gas option prevails, even with an optimistic 4 years life expectancy and subsequent disposal of the batteries.
This example applies to more than just chainsaws as it reflects all cost-efficient energy applications or simply where lightweight power and autonomy is required.
In other words, what works for forklifts does not apply to grids, cars or airplanes.
I think you will find that the biggest factor for battery powered carts and forklifts in industrial situations is the lack of emissions in an enclosed area.
Battery also acts as counterweight.
You completely overlooked the unobtainium-outworldium battery!
That’s the answer. /(sarc)
Otherwise, very instructive post. Thanks.
You are welcome. I really struggled to keep it simple. Because in fact it is complicated. Electrochemistry is a subject to which few aspire, since is a complex mix of physics and chemistry. Many of my examples are only ‘sort of’ true. Simplicity tradeoffs were made.
And for the record, I am self taught thanks to my issued patents.
Some of the units and discussion presented are just sloppy.
<i>”(1) energy density (how much charge they hold, aka how long they take to discharge, in Wh), more is better”</i>
Energy density is how much energy they hold PER CUBIC METER (or other favorite unit of volume. The units would be W⋅h/m^3 (or just J/m^3).
This is <i>related to</i> “much charge they hold” (presumable in A⋅h or perhaps charge density in A⋅h/m^3) and to “how long they take to discharge” (presumably in hr) but those are different concepts.
The units given — W⋅h — would be for total energy, not energy density.
<i>(2) power density (how many charge/discharge Amps per second), higher is better”</i>
This one is even sloppier.
Charge and discharge rates would be C/s or Amps, but certainly not A/s.
Power would be P = IV, measured in W or V⋅A. Both the amps and volts are needed!
Power density would be P/Vol in W/m^3
The same sloppiness continues later with “Second, flow batteries for the grid (where the charge is stored in the liquid electrolyte rather than in the electrodes, so with sufficiently big electrolyte tanks energy density is theoretically unlimited).”
It is the ENERGY that is potentially unlimited with an an unlimited size. The ENERGY DENSITY is set by the chemistry of the materials involved.
The idea that batteries are limited is important. The arguments would seem much more convincing if the basic physics was described correctly.
Terrific. Why don’t you contribute an intelligible comment correction here to my general posts, since you claim you are so ‘basic physics’ self proclaimed smarter in electrochemistry’, and so educationally superior.
I never claimed any of that.
Just to only have a few proven math theorem and issued patent insights.
Well, you DID claim that — I quoted you directly. Surely you don’t find the difference between energy and energy density or the difference between Amps and Amps/s unintelligible. You seem much too smart and informed for that.
The overall thrust of the article is interesting and informative. I enjoyed reading an learning a few new things.
Oh dear.
I’ll only bother to correct the first glaring error, Formula one engines do use lithium cells
Thanks Rud,
Interesting post and good comments.
Disappointed not to see a critique of cryogenic storage technology.
AFAIK these kind of systems can run at 60 to 65% round-trip efficiency, and more if there is a source of waste heat. They could also provide a source of coolth, so the economics would be interesting to explore.
I think you’ve been reading some highly misleading publicity. Here’s what the Jacobs report I cited says about LAES:
<blockquote>We have obtained estimated costs for LAES
systems from published figures provided by
Highview Power who recently commissioned a 5
MW pilot plant that had a stated capital cost of
£8m, with a round-trip efficiency of 25%. Highview
has recently announced plans for a 50 MW plant
with 5 hours of storage (250 MWh) and give
typical capital costs of approximately £1m per MW
installed for a plant with 4 hours of storage.
They also claim that a 200 MW plant with 10 hours
of storage could have a unit generation cost of
about £110/MWh generated, which implies a load
factor of 24% and assumes a round-trip efficiency
of 50%, although this is as yet unproven.</blockquote>
Highview claim higher efficiencies by assuming that “waste” heat or coolth (e.g. from LNG regasification) is outside the calculation. Of course, it has a cost of collection, and has to be included in the efficiency calculations. Otherwise you may be interested in a perpetual motion machine I have designed…
That hybrid of yours has a hitch. How’s the vehicle holding up under heavy towing loads, terrains, and weather?
Also., my all-gas 2005 Tahoe is easy to maintain and a lot cheaper to maintain than buying a new anything. Except for the “electricals” (circuits, switches, sensors, chips) being the growing problem with age. Downtime for repair far exceeds anything mechanical and often isn’t effective with the first or second try. ( I still consider even a hybrid to be early-adopter technology given the (vanishing) simplicity and reliability of what we already have.)
Just channelling ‘Limits to Growth’ – Peak lithium will be before peak oil/coal. Wshat then?
I have a relevant CFACT article coming out next week. NYC is bragging about building a 400 MWh battery system. They peak at around 32,000 MW and if they were 100% wind they need 7 days supply because they routinely get 7 day low wind heat waves. That works out to over 5 million MWh of storage needed to make 100% wind reliable. Oh and this is before Biden makes all the cars and trucks electric.
400 MWh is a very expensive meaningless toy. And utilities are playing this game across America, because the more they spend the more they make.
As I wrote almost a year ago:
https://www.cfact.org/2019/04/26/batteries-cannot-make-renewables-reliable/.
It cannot be done.
1) This source says NYC peaks around 13 GW, not 32 GW like you claim. This source is a few years old, so I can easily imagine 14 or 15 or 16, but not 32 GW. A smallish difference, but still a factor of 2x. Perhaps your number is for the whole state?
32 GW for 8 million people would be 4 kW per person, which is exceptionally high – even for peak demand.
http://www.peakpowerllc.com/notes/2015/2/17/whats-the-problem-with-peak-demand
2) Even if 32 GW is the peak, that only happens a few hours per day on the hottest days. Even on the hot days, the average might be 50% or 75% of that number.
3) No one is planning on 100% wind, so that is a complete red herring. NY also is planning a lot of solar (which peaks exactly during those hot summer afternoons) and NY currently has a lot of hydroelectric capacity (which can be ramped up at will). So only about 1/3 of the peak energy would need to come from batteries (the rest from hydro, solar, and a few conventional sources).
Put together, that means maybe 1/10 as much backup as you suggest to keep the lights on during that calm heat wave. Still a HUGE (and expensive) amount of back-up. But just using more realistic numbers has cut the project cost by 90%!
Also, I am sure you know that batteries are good for more than just back-up. They can provide local peak power, cutting down on the expense of additional peaker plants and upgraded transmission lines.
I’m sorry to say that this is the most poorly written and confusing post I’ve seen on this site. I’ve learned nothing.
if the messenger causes you to ignore the message you aren’t really listening … was it perfect ? of course not … but then again this is not an english comp or writing class paper being graded by an OCD professor … Is it ???
This (https://www.youtube.com/watch?v=NiRrvxjrJ1U) started out strong, then petered out toward the end. In fact, I didn’t finish watching it, even though the initial arguments seemed to make a lot of sense….
You wrote:
“We deliberately exclude indirectly kinetic pumped hydro, for which there are two big expansion problems: insufficient remaining suitable terrain, and California having rejected it, period.”
California!
There was some consideration of retrofitting Hoover/Boulder dam for pumped storage. Which seems a reasonable hope, technically anyway. Chemical batteries are a no-hope situation for utility-scale storage. “As you know, Russ.”
Calif has a number of fine candidates for pumped-storage at existing hydropower lakes.You want the high dams, and the highest (I think) is the one that had its spillway wash out in the last good wet winter. I can’t recall the name of it. Makes no sense to rule them out! Proven, fairly low-cost storage. Actually works, scales, and is well-understood. Then again, it makes even less sense to close Diablo Canyon, the last operating nuclear plant in the state. Weird place. Deliberate foot-shooting is a state specialty!