Re: Energy Storage method (2, try again)

7 views
Skip to first unread message

Keith Lofstrom

unread,
Mar 5, 2026, 9:40:50 PMMar 5
to Power Satellite Economics
> I would again like to focus on flow batteries using organic electrolytes from innexpensive materials available everywhere.

Are there any battery chemists on this email list who
have actually built a practical battery? Or are we
building castles in the clouds, expecting magic?

I created a NEW message, hoping to fork a discussion about
magnetic-levitated kinetic energy storage rings. That
message was immediately top-post-dogpiled with vague
non-expert battery talk.

----

I *HAVE* built meter-scale electronically-stabilized
magnetic levitators for steel bars; kinda spooky. Six of
those levitators didn't work. #7 was finicky. #8 is magic.

"Eighth try's a charm" is my exploratory engineering motto.

The bar "floats" below the electromagnets. Phototransistor
sensors "look" across the top of the bar and feed an analog
filter driving a power amplifier driving the magnets.
Nothing moving at high speed.

The top of the bar behaves as if rolling on frictionless
wheels over an invisible track. The levitator must be well
balanced, so the bar doesn't slide sideways; the residual
centering force comes from far-field effects on the ends.

Pulling down on the bar (by hand) is like pulling on a stiff
metal rail. A rail that slides sideways smoother than a
skater on ice.

Until the amplifier maxes out and breaks the feedback loop,
dropping and whacking the bar against my fingers over my
workbench. Ow. Ow.

----

With collaborators and a SAFE PLACE TO EXPERIMENT, I could
scale that bar 30x into a 10 meter diameter steel ring,
3 kg/m, deflected by magnets in a concrete trench.

With UNIFORM magnetic fields and control power similar to
my desktop levitator, magnetically deflected and NOT relying
on circumferential or radial mechanical strength, that ring
could rotate at 180 meters per second, 400 miles per hour.

Kinetic energy 1/2 * 3 kg/m * 30 m * (180 m/s)²
... approximately 1.5 Megajoules, 1.5 sticks dynamite

Hence the concrete trench and cover, because s**t happens.

Note that the ring is NOT under circumferential tension.
That might help with stability, or make it worse; I worry
about tensile stress waves and additional complexities.

I can fit that trench and ring in my back yard. My metal-
sculptor neighbor could help create it. Except that our
wives would divorce us, and the police would find laws
to prosecute us with. :-(

( I can imagine nasty consequences for a high energy
density flow battery, becoming an accidental blast
furnace instead. Better dying through chemistry. )

A 25 meter diameter circular trench with the same cross
section would also fit in my back yard, costing 2.5 times
as much.

What does 2.5x scale provide? The same V²/R centrifugal
force increases both R and V² by 2.5. For the same rotor
mass density (perhaps 3 kg/meter), the mass increases 2.5x
proportional to R. Stored energy increases by R*V² or 2.5²,
a factor of 6.25. Velocity increases by sqrt( 2.5 ) or
approximately a factor of 1.6, 640 miles per hour.

Stored kinetic energy increases to 9 MJoules, 9 dynamite sticks.

Except that the trench should probably be deeper, for safety.
Nonetheless, local police will probably call the FBI.
"Come out with your hands up and your machine STOPPED."
"But ... inertia ..." BLAM

So - let's find a location more permissive than suburban
Oregon. I ponder the vast fields of midwest agricultural
center-pivot irrigators, far below the Boeing 737 that
flies me east to visit my 106yo father-in-law in Maryland.
Those irrigator rings range up to 1600 meters diameter.

For a ring around a single 1600 meter diameter pivot field,
with the same "rotor" cross section, R and V² and cost
all increase by a factor of 160 over the 10 meter backyard
experiment. V increases to 2300 m/s, almost Mach 7 for an
atmospheric vehicle (which this is NOT).

Rotor mass, proportional to R, increases to 4800 kg.
0.5 M V² = 1.27e10 J = 12.7 GJ = 3.5 MWh .

Energy storage increases by a factor of 160 SQUARED for
160X cost. Much larger rings are possible, wider rotors
are also possible, and multiple rings can be nested
concentrically. All lowering the cost per stored KWh,
far below the cost of battery raw materials.

Again: Cost proportional to radius, stored energy
proportional to radius SQUARED. Chemical systems scale
with mass, at best. They self-ignite at worst.

----

But ... such enormous speeds! Won't residual gas drag
slow it down?

Consider the Large Hadron Collider, which moves protons
at 0.99999999c, 2 meters per second slower than the speed
of light, Mach 875000 ... for HOURS without collisions
in a very very VERY good cryo-pumped vacuum.

A solid ring moving at 2300 m/s in a "merely good" vacuum
endures practically zero drag, with a "noseless" toroidal
rotor configuration and commercial grade vacuum.

Control signals in optical fiber move 200,000 km/s, and
5 kilometers of armored single mode fiber costs $10K.
Compared to electronic and signalling speeds, the ring
is barely moving.

----

What if: Larger than that? FASTER than that?

Too dangerous on land, with people living nearby.
Instead, rings in waterproof shells, submerged in tens
of meters of seawater, can grow to hundreds or thousands
of kilometers circumference, and move at Earth orbit
speed or faster. Earth orbit speed allows the rotor to
follow the curve of the Earth in a "zero gee ballistic"
with less magnetic deflection.

Magnetic CORRECTION *ALWAYS* required, but that is
"speed of electronics", vastly faster than orbital speed.

When a power storage ring fails deep in the ocean, there will
be a LOT of boiled seawater and vaporized rotor fragments.
Soon after, plankton blooms growing in the iron-enriched
seawater, feeding extra fish.

Not much "extra" on the global scale, but not a disaster
either, given adequate preparation and forethought.

Forethought is my main concern. Scum floats to the top
of most human organizations. The same psychopaths who
brought us Pearl Harbor, nuclear weapons, Challenger,
Chernobyl, and countless "Wars to End All Wars" will shave
safety budgets into their own pockets, hamstringing
ethical engineers.

So, please help me design fail-safes with redundant
fail-safes. Think twice before you react disparagingly.
You might think of a valuable improvement instead.

And after all that, PLEASE read http://launchloop.com/PowerLoop

... then THINK about it. "L5 in 95" didn't happen because
we didn't think ENOUGH.

----

If you read down this far, congratulations, you are more
patient and thoughtful than a world of lazy hate-tweeters.
Please help me keep THIS thread on track; summarize and
BOTTOM POST. We won't reach the stars by chasing our tails.

Keith L.

--
Keith Lofstrom kei...@keithl.com

Simon Quellen Field AB6NY

unread,
Mar 6, 2026, 9:12:03 AMMar 6
to Keith Lofstrom, Power Satellite Economics
Let's look at the 800-meter radius ring spinning at 2300 meters per second, weighing 15,080 kg.
That's a hundred million Newtons (99.7 million).
That's 10,000 metric tons of outward force.

If that isn't countered by the tensile strength of the material, it has to be countered by some very large superconducting electromagnets. The cryogenic cooling costs have to be calculated, as well as the amortized cost of the plant and equipment over the lifetime of the project, and the interest on the debt to finance it.

At Mach 6.7, the aerodynamic drag at 10e-3 Torr is not zero. The remaining air would be a plasma, melting the ring. To prevent this, you would need an ultra-high vacuum (10e-7 Torr), which requires continuous pumping, an energy cost not accounted for in your analysis.

I assume you are using a linear synchronous motor to charge and discharge the device. These are 85% to 92% efficient. That loss adds to the losses from vacuum pumping and cryogenic cooling of the magnets. The total efficiency is likely to be about 50% after these losses. Compare that to the 90% of batteries, 85% of traditional flywheels, or 70% of pumped hydro.

If you think I have not thought twice before disparaging the idea, consider that the 15,080 kg mass I used was arrived at by using the circumference of your 800-meter ring instead of the diameter, which you mistakenly used. The stored energy is 11 MWh, not 3.5 -- a factor of three greater.

Always good to read your stuff, even when I am disparaging an idea.
:-)



Virus-free.www.avast.com

--
You received this message because you are subscribed to the Google Groups "Power Satellite Economics" group.
To unsubscribe from this group and stop receiving emails from it, send an email to power-satellite-ec...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/power-satellite-economics/20260306024043.GA13263%40gate.kl-ic.com.

Geoffrey Landis

unread,
Mar 6, 2026, 3:38:28 PMMar 6
to Power Satellite Economics
>Kinetic energy 1/2 * 3 kg/m * 30 m * (180 m/s)²  ... approximately 1.5 Megajoules, 1.5 sticks dynamite

Sounds like a lot, but that's only 0.4 kW hours. With the US average cost of electricity currently 17 cents per kW-hr,  your 33-foot diameter ring stores seven cents worth of energy.
The 82-foot diameter scaled-up version stores about 42 cents worth of electricity,  I'm doubtful that it's cost competitive.

>Note that the ring is NOT under circumferential tension.

I'm missing something. Why wouldn't a spinning ring be under circumferential tension due to centrifugal force?

--Geoff

Paul Werbos

unread,
Mar 6, 2026, 3:53:16 PMMar 6
to Keith Lofstrom, Power Satellite Economics, Vidvuds Beldavs

On another list, I commented on the use of flow batteries, which were well within my scope when I funded electric power at NSF.
Here is part of that other discussion:
Good morning, Alejandro!

Thank you again for the discussion yesterday. I am glad to hear you will be visiting China again soon. I should inform the others...
Alejandro is a crucial, central person in the Solar Energy Research Center/Consortium of Chile (SERC), which in my view is right at the center of humanity's best hope of avoiding extinction by climate change in this century. I hope that there will be enough useful information in what I send you to fully justify your time, and excuse whatever part of my emails is less useful.

Yesterday I mentioned the two top people in real RLADP in China, but I didn't immediately remember then name of the first modern battery manufacturer
I met. In a higher state of consciousness this morning, I CERTAINLY remember Winston Chang of Thunder Sky, the first manufacturer of the breakthrough batteries which have stimulated good competition and growth and so on. In a way, half my present family income is due to meeting him!

I met him in Shenzhen, in his company building, including the manufacturing and test facilities. He had a great woman I liked representing and explaining the company! Thanks to Sam Leven, a group of us later arranged for a lawyer named Vergil to follow up on that visit; he liked her even more, married her, and moved to China. (I do not know where he is now. It is conceivable Sam or Rodney Smith might know, but I do not even have Rod's new email.)

Most stunning to me were the test data on new batteries with about 150,000 miles driving range, for normal car batteries and very quick recharge.
I was actually chosen to be the main speaker presenting George W. Bush's "Energy Independence and Security Act" (ACT) to big plenary talks at the 
US Congress (House and Senate, including the Rayburn Foyer which held about 200 people); I still have the slides, which almost certainly include Thunder Sky,'like many talks I gave back then.

In 2009, my year of working as a Fellow in the US Senate, I published an IEEE paper on the use of new, RLADP control for battery management.
That became the start of a small company set up by Rod, Sam and my wife, which was later sold to Wartsila. The Wartsila people never mastered that technology, in part because my duties at Senate and NSF did not give me time. HOwever, my duties at NSF DID give me time to organize
the best workshop on advanced batteries every held then, at MIT, under Sadoway. The Thunder Sky battery worked well, of course, but the best by far was a rechargeable lithium-air battery developed by Lonnie JOhnson of Atlanta. Argonne National Laboratories verified it would continue working after many, many recharges; it would have 5-10 times the weight to kwh as the best current batteries (like Thunder Sky and BYD and CATL). At NSF I was able to fund the initial work, but was VERY deeply horrified by the forces of corruption I saw in the US which diverted US funding to big rich companies TRYING (but failing) to reinvent what Lonnie had already done. This is not the most important opportunity for SERC involvement, because the other opportunities are even more urgent, but in a rational world money should be available somehow for all of these. A faster growth of electric cars, trucks and airplanes (4000 km range available for 777 class aircraft) would help our chances of climate survival and even expand world trade. 

Also in 2009... when I saw that Obama's draft climate bill was a total loser (for many reasons I would be happy to discuss if anyone is interested),
Rod and Sam helped contact Jack Lin Tang, and all helped create a meeting between Biden and Hu JinTao which made many newspapers back then. 
With our support and inputs, Hu proposed an alternative climate agreement, far more rational, efficient and market-based than Obama's bill.
I was deeply disgusted that Obama listened to Biden when he and Reid told Obama they depended on deals that had been made with certain stakeholder friends, and that they could not allow an alternative. BUT NOW WHEN HU MAY BE CLOSER TO XI'S NEW POWER CENTER... 
THERE MAY BE ADDITIONAL CRUCIAL NEW OPPORTUNITIES COMING OR CREATABLE, IF ALL REMAIN VERY ALERT TO ANY NEW HINTS.

Alejandro Jofré ajo...@uchile.cl

Fri, Sep 5, 2025, 1:07 AM
to ilinexjacklintangsamnetsme
Thank you,  Paul, for your message containing a lot of information!

During my visit to China, starting September 21, we will stay a couple of days visiting CATL’s labs in Nigde. We are starting a cooperation with them. They have inaugurated offices in Santiago this week, not far from the SERC meeting. After we visit China, we will spend another week in Japan. I will come back to Chile on October 5th. So, we can then discuss the next steps. We could organize a second Zoom the week of October 6th.

Best regards,  Alejandro
-----

=====================\=======\======
============================

P.S. About half my family income comes from sale of a small business I started to use the Chinese batteries 
in electric power demand management. The VCs went quick and dirty, just as Musk did years later when he aimed at the exact same market niche,
but we sold our rights before that.

Robert Poor

unread,
Mar 6, 2026, 4:09:50 PMMar 6
to Paul Werbos, Keith Lofstrom, Power Satellite Economics, Vidvuds Beldavs
Keith:

I'm sorry that your PowerLoop proposal got stepped on by battery discussions -- it warrants unencumbered debate.  And I'm glad to have seen the replies from Simon and Geoffrey -- they helped me understand some of the challenges.  I won't debate the technical merits of PowerLoop (except to say that the failure modes sound, um, exciting), but I will fall back to my previously stated position that:

"The storage technology most likely to become the dominant standard is one that gets cheaper over time and exhibits fast learning curves."

While PowerLoop may succeed on the first requirement, it's hard to imagine learning curves fast enough for it to become economically viable.  (As an example of a technology that fits both requirements is chemical batteries, which has shown a 10x reduction in $/KWh since 2010 and for which there's a continuous deluge of press releases about improved chemistries, solid-state batteries, sodium-ion batteries, etc. But I said I wouldn't hijack the PowerLoop thread, so I'll stop.)

- rdp

Keith Henson

unread,
Mar 6, 2026, 6:41:31 PMMar 6
to Geoffrey Landis, Power Satellite Economics
On Fri, Mar 6, 2026 at 12:38 PM Geoffrey Landis
<geoffre...@gmail.com> wrote:
>
> >Kinetic energy 1/2 * 3 kg/m * 30 m * (180 m/s)² ... approximately 1.5 Megajoules, 1.5 sticks dynamite
>
> Sounds like a lot, but that's only 0.4 kW hours. With the US average cost of electricity currently 17 cents per kW-hr, your 33-foot diameter ring stores seven cents worth of energy.
> The 82-foot diameter scaled-up version stores about 42 cents worth of electricity, I'm doubtful that it's cost competitive.
>
> >Note that the ring is NOT under circumferential tension.
>
> I'm missing something. Why wouldn't a spinning ring be under circumferential tension due to centrifugal force?

It is supported by superconducting bearings.

KeithH
> --
> You received this message because you are subscribed to the Google Groups "Power Satellite Economics" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to power-satellite-ec...@googlegroups.com.
> To view this discussion visit https://groups.google.com/d/msgid/power-satellite-economics/343a24a8-9f11-446e-8536-b64c74e5e285n%40googlegroups.com.

Simon Quellen Field AB6NY

unread,
Mar 6, 2026, 9:00:55 PMMar 6
to Keith Henson, Geoffrey Landis, Power Satellite Economics
Let's look at those bearings.
The ring spinning at 2300 meters per second feels a force of 19,837.5 Newtons per meter trying to tear it apart.
Any magnetic force trying to contain that will experience the same force in the opposite direction.
Magnetic force drops off quickly with distance, so you will need to have the magnets very close to the ring.

I don't see any calculations for the hysteresis forces or the eddy current forces that both reduce the efficiency and heat the ring. Cooling the ring in a vacuum will be difficult, and you will need to keep the temperature below the Curie point.

The need to avoid the Curie point will limit the power you can put in or take out. In other words, the rate at which you can add or remove energy will be limited.

Something is going to have to handle the 674 G of acceleration. If it isn't the ring itself (steel can handle 2 GPa at best, not the 41 GPa in this setup), it is the magnet holding it together.

I've been assuming a magnet located slightly above the center of the ring, providing both lift and centripetal force. The lift force would be 148,000 Newtons (15,078 kg * G). The force holding it together would be 674 times as great (roughly 100 million Newtons).

Replacing the central magnet with a ring of magnets just barely inside the steel ring would probably be best. But it would still have to resist 10,000 tons of force trying to rip it apart.

Magnetic force from a dipole drops off with the cube of the distance. The inside of the ring will see more force than the outside. 10,000 tons of force would be over 250 GPa, and steel can only take 2 GPa, so the magnetic force can't drop much across the 20 mm thickness of the bar before we get explosive disassembly.

The approximately 200,000 liters of liquid nitrogen needed to cool a 5 km magnet ring would require between 5 to 10 megawatts of power to cool. The 5 km magnet ring will contract by as much as 20 meters while cooling from ambient to 77 K, so it will have to be designed for that, to the same tolerances we are assuming for the separation between the magnets and the steel ring.


Virus-free.www.avast.com

Bryan Zetlen

unread,
Mar 6, 2026, 9:35:19 PMMar 6
to Simon Quellen Field AB6NY, Keith Henson, Geoffrey Landis, Power Satellite Economics
There are those times when it’s just plain entertaining to hear about arcane Rube Goldberg ‘disruptive technologies’, ‘advanced storage schemes’, ‘global scaling’ yada yada. Flywheel energy storage was first proposed in the late ‘50s, early ‘60s using electric locomotive trains with 100s of rail cards loaded with concrete on velodrome banked raceways. These closed systems would use surplus hydro generation to accelerate to 100+ mph. Then being drawn down during peak demand. One of my responsibilities at Boeing Phantom works was structural analysis on Boeing’s cryogenic flywheels using superconductive ceramic flywheels. Folks ain’t nothing new under the sun. 

Roger Arnold

unread,
Mar 6, 2026, 9:46:28 PMMar 6
to Keith Henson, Geoffrey Landis, Power Satellite Economics
Oh, why not? Since everybody's piling on here, I guess I'll join the festivities...

3 kg/m is a ridiculously low mass for one to use if the intent is to design a serious energy storage device. I think that that was just for a ring of single rebar that Keith L. was using to illustrate the possibility of energy storage in a magnetically levitated ring. Anyhoo,.. It turns out that mass per meter is not what one should be looking at to calculate the energy storage capacity. What counts is the average flux density of the levitating magnetic field. The same levitating field can support a light ring spinning fast, or a heavier ring spinning slower.

Without diving into the details, it also turns out that one can get a. pretty decent average flux density for the levitating field without resorting to superconductors. One Tesla can be managed with NdFeB permanent magnets attracting soft iron. I believe Keith L. was thinking in terms of variable current-controlled electromagnets for levitation, and that's indeed how the Trans-rapid Maglev train system works. But the variable current continuously flowing through the copper coils of the electromagnets creates ohmic losses. The losses are quite tolerable for a maglev train. I believe they correspond to a coefficient of friction of about 0.1% when the train is running at speed. But they make that suspension system unsuited for energy storage for any period exceeding a minute or so. For long term energy storage you need something better.

I have what I think is a pretty slick design for that "something better", but I won't go into that here. It's not that I'm trying to protect a design that I expect to patent, it's just that it's too hard to explain it in a post like this. But I'll give a hint: it depends, in part, on switchable permanent magnets. I.e., permanent magnets that can be rapidly turned on and off. The bottom line for energy storage capacity in a 10 m diameter, 2 m tall cylindrical flywheel is ~63 MJ.About 17.5 kWh. But the capacity scales with the cylinder height times the square of its diameter. I.e., the volume of the space enclosed by the cylinder. 

If anyone wants details, they should let me know. I haven't written the tech paper yet, but I expect I'll be doing so soon-ish. My e-mail is silver...@gmail.com. (Same as on the distribution list for this group).

Roger

On Fri, Mar 6, 2026 at 3:41 PM Keith Henson <hkeith...@gmail.com> wrote:

Simon Quellen Field AB6NY

unread,
Mar 7, 2026, 9:10:59 AM (14 days ago) Mar 7
to Keith Lofstrom, Power Satellite Economics
You are backing off from the 2300 meter per second speed to 693 (Mach 2).
The hoop stress is now 3.77 GPa, which is very close to the limit of what steel can tolerate.

The negative stability is still troublesome. At that speed, the band moves 0.7 mm every millisecond.
While the computer has plenty of time to calculate the adjustment, you are trying to make a half Tesla magnet change at that speed. The time constant of the magnet will be single-digit milliseconds, and the ring will move 70 centimeters in a millisecond. You will likely need a gap of at least 5 cm to react fast enough. Increasing the gap means increasing the magnetic field and the current in the coils.

This is a big gyroscope. The angular momentum is enormous. A small earthquake moves the ground a couple of centimeters at a frequency in single-digit Hertz. The magnet needs to be strong enough to correct for that movement before the ring crashes into something, despite the gyroscopic forces countering the correction. That may be far more magnetism than you are counting on for levitation and cohesion.

You still have not addressed eddy current heating that might exceed the Curie point (770 C).
If the temperature rises by even 50 Celsius, the ring expands by 3 meters.
The air friction at 0.1 Torr will still cost 100 kW to overcome, and generate heat that is hard to get rid of in a vaccuum. The system will self-discharge at 4% to 6% per hour due to air friction.
Mainenance on a 0.1 Torr vaccuum with this many joints, sensors, and power feedthroughs will be a constant headache.
What is the acoustic resonance of the ring? Will the magnetic pulses ever hit that frequency, causing the ring to vibrate like a guitar string and snap?
The safety casement needed to withstand 700 m/s steel is 2 meter-thick concrete. The 5 km circumference will require 100,000 meters of concrete, or excavation of 80,000 cubic meters of earth. The concrete would cost $150 million, but the excavation would be a tenth of that.
However, excavation requires continually pumping out ground water, and makes removing the 100 kW of air friction heat even harder. You'd need refigeration or a water jacket around the 5 km ring. The heating would raise the soil temperature by 2 Celsius every day, reaching the boiling point in a month, drying out the soil and making it even more of an insulator.

By reducing the speed from 2300 m/s to 700 m/s, you have increased the feasibility.
But there are still many problems you have not addressed.
Some of these may be mitigated by increasing the scale, but that makes others even harder (such as maintaining the vaccuum).



Virus-free.www.avast.com

On Sat, Mar 7, 2026 at 4:13 AM Keith Lofstrom <kei...@keithl.com> wrote:
On Fri, Mar 06, 2026 at 06:09:37AM -0800, Simon Quellen Field AB6NY wrote:
> Let's look at the 800-meter radius ring spinning at 2300 meters per second,
> weighing 15,080 kg.
> That's a hundred million Newtons (99.7 million).
> That's 10,000 metric tons of outward force.
>
> If that isn't countered by the tensile strength of the material, it has to
> be countered by some very large superconducting electromagnets.

I may have been unclear.
Let me explain in "moves-lips-while-reading" detail.

Superconductors are superfluous.  Sub-Tesla copper and
transformer steel magnet are more than adequate.

(note: My 1975 UC Berkeley MSEE thesis involved
superconducting circuits.  That was HARD.  Colleagues in
the UCB physics department won the 2025 Nobel Physics prize
for related work ... and they deserved to win it in 1985).

----

(1) The rotor is NOT circle-cross-section toroidal, instead
a "flat" band/cylinder of iron, "magnetically" resembling
"unwrapped" the outer surface of a multipole non-synchronous
induction motor.   

Multiple circumferential alternating N/S pole BANDS, on the
order of 10 cm spacing.  PERHAPS with windings and optically
triggered triacs to "lock in" a induced field.  That might
slightly improve efficiency, but adds cost and complexity
and reduces reliability.

The rotor band is magnetically deflected, and NOT under
strong longitudinal tension; just enough tension to keep
it "predictable" and "straight enough" to correct spacing
electronically.  There is probably a sweet spot for that,
because magnetic deflection of a ring can have negative
stability without active predictive/adaptive control.

The system is stabilized electronically and computationally,
with electronic signals moving at fractional light speed.
Compared to the electronics, the mechanical velocities
are tiny.  System clock rates on the order of GHz; highly
parallel GPUs clock hundreds of cores at 2 GHz today.

Some WAG numbers, NOT OPTIMIZED:

(1a) magnetic attraction "pressure" scales as B²/2μ₀.
For a 1 Tesla field, that is 1²/(2*4e-7*π) ≈ 400K Pa
or 40 Newtons per square centimeter.  With multiple
"pole bands", field spreading, loose geometry, and
Murphy's Law, call that 5 Newtons per square centimeter
averaged over the width of the rotor band.

( Superconductor magnets can do 30 Teslas, 900 times the
magnetic pressure, but they become NON-superconductors
if you look at them funny.  vini, vidi, v×c=q.  I came,
I saw, I have been conquered by stray capacitance. )

4 tracks of 5 newtons per square centimeter across a 6 cm
wide, 1 cm thick iron rotor band is 3000 newtons per meter.
If the rotor band is 5 mm thick, 8g/cm² iron, it masses
approximately 5 kg per meter.  The centripedal acceleration
on the curved band from curved magnets is 600 m/s², 60 gees. 

For a circular band and magnet system 5 meters radius,
v² = a × r = 3000 m²/s² so v can be 55 m/s .  The band
circumference is 10π meters (round that to 30 meters)
so the mass is 150 kg. 

Total kinetic energy in this "small" rotating ring is
E(5m) = ½ × 150 kg × 3000 m²/s² = 225,000 Joules

For an unchanging system, the B field is constant,
dB/dt is zero, no induced voltages in any conductors
(unless you try to steer it fast). 

----

Now, double the radius, pretty much doubling the cost
of materials, trenching, etc.

r doubles, v² doubles, mass doubles.
E(10m) = ½ × 300 kg × 6000 m²/s² = 900,000 Joules

Doubled cost, quadrupled stored energy.  EXCELLENT
scaling behavior.

----

For an 800 meter radius "around a pivot irrigator" ring

E(800m) = ½ × 24000 kg × 480000 m²/s² = 5.76e9 Joules
                                      = 1.6 MWh

80x more cost, 6400x stored energy.

SQUARE LAW value versus cost.  You don't find bargains
like that every day ... unless you are an integrated
circuit designer like me.

In Real Life, the storage SYSTEM will probably be a race
track oval with straight sections passing through for
linear induction motors.  The hypothetical "sweet spot"
is "straight sections" following the curve of the Earth
at orbital velocity. 

Maximum value if the loop system spans the Pacific ocean,
moving energy from summer to winter, daytime to night ...
but let's save that for later, and focus on inexpensive
prototypes and introductory products. 

-----

I omit the details of transferring power into and out of
the ring, but I have a shelf of linear motor engineering
books, and can suggest (different) books for others with
weak or strong math backgrounds. 

The 1600 meter circumference ring velocity is 700 meters per
second - it must be in a decent vacuum.  An ultra-high vacuum
like the "speed of light" Large Hadron Collider is NOT required. 

However, for very large rings with VERY high velocities, even
in vacuum, loose chunks can trigger "hypervelocity spalling
cascades".  Special coatings and fragment diverters required.
Lots of learning ahead, LHC wasn't built in a day.

----

Someday:

An enclosed ring moving faster than orbital velocity can
circle the Earth and support cables to the ground.  For
an electronics designer like me, Earth orbit velocity
is 8 micrometers per nanosecond. 

Paul Birch (r.i.p.) wrote about rings in JBIS, and Ken
Brakke wrote about orbital rings in Analog, one month
before my launch loop article in Analog, more than 40
years ago.  We collaborated, and standardized nomenclature.

I may still have the PDFs for those articles on my
launchloop.com website.  Paul was Cambridge, Ken was
Princeton, and both were far better mathematicians than me.

----

BTW, Ken lives near Bellingham Washington.  We discuss
space solar power USED IN SPACE for Bitcoin mining and AI.
Don't move space gigawatts ... move milliwatts of encrypted
data, while radiating waste heat into the 2.7K void.  High
energy radiation?  That's how we MAKE chips these days.

----

In summary: The point of my writing is expressly the OPPOSITE
of exotic technology; instead, it is about the amazing "cost
squared" performance of low-tension magnetically-deflected
iron rings.  Combine that with Moore's Law ("not dead yet"),
then unlock the solar system.

Keith L.

P.S.  Let me repeat:


> > If you read down this far, congratulations, you are more
> > patient and thoughtful than a world of lazy hate-tweeters.
> > Please help me keep THIS thread on track; summarize and
> > BOTTOM POST.  We won't reach the stars by chasing our tails.

Why?  Almost nobody reads what's below a top post.  Sadly,
in 2026 almost nobody reads anything longer than a tweet.
Ever seen a tweet with calculus?  Chinese do not top post ...
and China is cleaning our clock technologically.

or in "ultimate top posting" format:

technologically clock our cleaning is China, and ... post
top Chinese not do.  calculus with tweet a seen Ever?
tweet than a longer anything reads almost nobody 2026 in,
Sadly.  post top a below what's reads nobody Almost.  Why?

--
Keith Lofstrom          kei...@keithl.com

Roger Arnold

unread,
Mar 7, 2026, 7:34:40 PM (13 days ago) Mar 7
to Power Satellite Economics, Keith Lofstrom, Simon Quellen Field AB6NY
I don't know if my previous reply on this thread got lost, but I'll reiterate part of what I said there;

Both the mass per unit length and the rotational velocity of the flywheel ring are -- I can't say "irrelevant", but "secondary" perhaps -- to the energy storage capacity of the ring. The primary determinants of capacity are the diameter and height of the flywheel ring and the strength of the levitation field that resists the centrifugal force of the spinning flywheel. The levitating field can be either attractive (pulling the ring toward the axis) or repulsive (pushing it away from the outer containment wall toward the axis). An average field strength of 1 T (Tesla) corresponds to a field pressure of + (repulsive) or - (attractive) 58 PSI. 

There's little if any reason for the ring to spin faster than what will generate a centrifugal force of ~10 G. At 10 G, 9% of the levitating force will go to countering the gravitational weight or the ring, while 91% will go to countering centrifugal force. If it were somehow important to reduce the mass of the ring, then it could be thinner and spin faster, but the storage capacity would remain set by the strength of the levitating field.

The key feature of this type of energy storage is the way it scales. The cost scales with the circumference of the evacuated tunnel in which the ring spins. The energy scales with the area enclosed. So no matter how high the cost per meter of ring and tunnel turns out to be, there's a radius at which the cost per kWh is arbitrarily low. But there has to be a market for the resulting energy storage capacity.

I was calling this type of energy storage ring an "externally contained flywheel" when Keith Lofstrom and I discussed it. That was sometime back around the dawn of history.  Or late in 1979, actually. I remember, because it was just after "The Spaceport" was published in Analog. Keith read it, found my phone number, and drove up to visit me in Washington. I was working for Boeing Aerospace at the time, and was president of the NW Chapter of L5 Society. Fun times. I didn't push the idea then, because intermittent renewables were not yet a thing. There was simply no market for the scale of energy storage that would have been needed to justify the development.

Oh, a final note: keeping the ring cold is no problem, because there are no hysteresis or eddy current losses in the ring itself. It can be made a passive and non-conducting component. There are certainly non-zero losses in the system as a whole, but they arise almost entirely within stationary components along the walls of the enclosing tunnel. The losses are small, and the components can be actively cooled. The only heat generated in the ring is frictional, from residual atmosphere within the tunnel. That can be made as low as needed by attention to leakage into the tunnel and the quality of its vacuum pumps.

- Roger

--
You received this message because you are subscribed to the Google Groups "Power Satellite Economics" group.
To unsubscribe from this group and stop receiving emails from it, send an email to power-satellite-ec...@googlegroups.com.

Tim Cash

unread,
Mar 10, 2026, 9:15:51 AM (11 days ago) Mar 10
to Roger Arnold, Power Satellite Economics, Keith Lofstrom, Simon Quellen Field AB6NY
I am curious about how these numbers and equations were derived: Was it using the human noggin without aid of any online AI tools, was it a collaboration of both, or was it strictly AI?
I say this because I have been "running a problem" to find out if there is in fact, an "optimum altitude" above the ground to launch a rocket/upper stage from into orbit that saves energy from air drag as compared to launch from the ground?  I have "invested" many evenings using AI only just to have the Answers chasing the Questions tails, and ending up with a Mess that cannot be sorted out.  A much better way to solve any problem of this magnitude is to lay it out before you with pencil and paper before committing it to computer code for solution, and perhaps the next step would be to build test models and fly them to find a solution to the problem?  In any case, I simply do not and will not trust any answers that I do not have to "pay for".  It does not make sense to me that anything of value is offered "free as in Beer" for the asking.  However, if we are searching for the mythical ultimate energy storage solution, would not a black hole qualify for that award?  It may present a problem to anyone trying to tap into that energy sink without falling into the black hole, and getting "spaghettified" by the process?  Does it not make sense that if there were an "optimum altitude" for air launch, that more than just a few aircraft launch platforms would have been tried by now?  It does to me.

Regards,

Tim Cash



--

Roger Arnold

unread,
Mar 10, 2026, 2:43:22 PM (10 days ago) Mar 10
to Tim Cash, Power Satellite Economics, Keith Lofstrom, Simon Quellen Field AB6NY
Hi Tim,

I can't speak for Keith or Simon, but personally, I'm too much of a dinosaur to be using AI as anything more than a smart search engine and calculator. I wouldn't know how to use it for anything as complex as what you're trying to tackle, and I wouldn't trust the results if I figured out how to try.

My recommendation for limiting centripetal acceleration in the energy storage ring to no more than 10 G wasn't the result of any hard optimization analysis. I simply assumed that a major cost component for the system would be the aggregate levitation force. The levitation force has two parts: one supporting the mass of the ring against gravity, and the other supporting it against centrifugal force. Energy storage capacity is proportional to the part of the levitation force acting against centripetal acceleration -- the part that keeps the ring from crashing against the wall of its containment tunnel. The part that acts against gravitational acceleration is non-productive overhead, to keep the ring from crashing against the floor of its tunnel. 

At a centripetal acceleration of 10 G, we have 10 parts of levitation force containing the ring and contributing to its energy storage capacity, and one part supporting it against gravity and not contributing to its energy storage capacity. That's 91% of the levitation force going for energy storage capacity, vs. 9% for the unproductive overhead of keeping the ring levitated. Yes, we could raise that 91/9 ratio by designing for a higher ring velocity / higher centripetal acceleration, but we're already into a region of diminishing returns. And a higher ring velocity creates various complications and costs that likely outweigh any marginal gain from a higher fractional utilization of the levitation force.

The problem you're wrestling with is far more complicated. You're really dealing in large part with issues of aircraft design. There are sharp discontinuities in the design space depending on whether the launching aircraft is subsonic, supersonic, or hypersonic. The relatively high aspect ratio wings of subsonic aircraft deliver vastly greater lift to fuel consumption ratios than any alternatives, but come with vastly lower ratios of lift to mass. They're very fuel-efficient, but very heavy with respect to the lift they can deliver. At the other extreme are rocket engines. They guzzle fuel (reaction mass) like crazy, but deliver fantastically high lift to mass ratios. That's why rocket designs have mostly stayed with boosters that take the delta V hit of a low vertical liftoff acceleration and high gravity losses, compared to what could theoretically be achieved with HTOL aircraft boosters. The weight penalties and design complications that come with an HTOL design outweigh the savings in delta V to orbit that come from a more efficient climb through the atmosphere. It's easier just to make the fuel tanks larger and burn through the extra fuel that VTOL requires.

Complexities notwithstanding, best of luck in your explorations.

Roger
Reply all
Reply to author
Forward
0 new messages