Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Capture energy from lightning

  1. Feb 25, 2004 #1
    two very large solenoids (tons maybe ) Stacked vertically, very large capacity wire coil w/ iron core. upper solenoid attached to guide rails to keep its trajectory vertical.
    now introduce current I.E. lightning strike. how can one tell if the upper solenoid will be lifted or how far it would be separated from the lower?
  2. jcsd
  3. Feb 25, 2004 #2


    User Avatar

    Staff: Mentor

    The difficulty in mechanical capture of the energy in lightning is that there is a lot of energy in an extremely short amount of time. That's not much time to charge an inductor or accelerate a mass.
  4. Feb 26, 2004 #3
    envision a hollow tower w/ a leading edge technology wind turbine @ the top, leaving the tower a lightning rod mast , inside the tower the solenoids, one on guide rails extending to the top of the tower.
    the lightning rods insulated ground w/ appropriate gage wire cable runs down the tower and is distributed to the solenoids.
    what gage is the cable in order to conduct w/out melt down? what weight is the solenoid & what will limits its travel up the tower other than gravity and friction (the extent of the magnetic field?) ,
    how tall is the tower that will contain the solenoid / projectile?
    what to do with the stored energy at the top of the tower when the weight is at its apex would be the next matter to consider.

    The difficulty in mechanical capture of the energy in lightning is that there is a lot of energy in an extremely short amount of time. That's not much time to charge an inductor or accelerate a mass.

    would the size of the wire help here?, or some sort of buffer?
  5. Feb 27, 2004 #4


    User Avatar
    Science Advisor

    First off would be how quickly the charge transfer occurs with a lightning bolt. One discharge lasts only 60uSec according to the link below.


    In addition, note the air is heated to 50K F. That's where most of the work is going is into heating the air. Read near the bottom of this link, it references the same book and provides more info on how impractical lightning capture is as a power source.


  6. Feb 28, 2004 #5
    First off would be how quickly the charge transfer occurs with a lightning bolt. One discharge lasts only 60uSec according to the link below.

    would rapid charge transfer cause explosive acceleration of the solenoid in our hypothetical tower ?
    what happens to all that heat when lightning strikes a lightning rod? If the lightning rod won't melt there shoudn't be a problem w/ the conductor to the

    FROM CLIFF_J Read near the bottom of this link, it references the same book and provides more info on how impractical lightning capture is as a power source.

    I will look at the library that sounds like a good reference book, meanwhile to make it practical is an interesting puzzle, besides wasn't it fun when you did something they said you couldn't ?

    Thanks for the links, very interesting stuff.
    The average lightning stroke has a peak electrical current of 30,000 amps. (taken from another link ; The average lightning strike contains 20,000 amps.)

    The average flash will light a 100 watt bulb for more than 3 months.

    Lightning’s heat exceeds 50,000 degrees F. or three times hotter than the surface of the sun.

    Electrical current travels along the outside of a conductor, as in a person struck.
    "In reality, lightning often flashes over the outside of a victim, sometimes blowing off the clothes but leaving few external signs of injury and few, if any, burns".

    Uman, M "All About Lightning", Dover, 1986.

    Rakov, V. A. and Uman, M. A: "Lightning: Physics and Effects." Cambridge University Press, 2001.

    Source: USA TODAY research by Chris Cappella
  7. Feb 28, 2004 #6


    User Avatar
    Science Advisor

    Your biggest problem is likely the inductance of the wire itself as a straight piece of wire, much less the wire wound in a coil.

    Any impedance means the current will be converted into heat. Current squared times impedance tells you the heat generated, and this would be a lot. Might need like MCM500 wire or something...

    Its always nice to prove something can be done when everyone says it can't. I was pretty sure I could trisect an angle at one point! :) I think we'd all like for you to prove us wrong and provide a meaningful alternative energy source (any of us that have been to the gas pump lately anyways) but you do have convential wisdom working against you.

  8. Mar 1, 2004 #7
    Cliff- Your biggest problem is likely the inductance of the wire itself as a straight piece of wire, much less the wire wound in a coil.

    Avoiding the heat problem seems to be a simple problem unless I am missing something ( wouldn't be the first time). The questions I can't answer myself are like these.
    will the solenoid accelerate to rapidly (or explode)? How would one arrive @ the optimum weight of the solenoids (goal being to lift the greatest weight possible under control). Any help is appreciated, thanks for your input Cliff.

    From; www.lightningrodparts.com/parts1.html
    - Standard of the industry for structures less than 75' high.
    Catalog Number 1 - 32 strands of 17 gauge bare copper wire in smooth or basket weave configuration. 7/16" diameter, 65,500 circular mil cross section. Weight of copper - 204 lbs. per 1,000' - (Other size cables are available)
  9. Mar 1, 2004 #8


    User Avatar
    Science Advisor

    Lets pretend the wire can capture the energy without losses to heating the air to 50,000 F. Inductance is the resistance to change in current. You're asking the wire to flow 10-30kA (very tough job) with a rise time of like 10uSec or 10KHz.

    Here's a formula for straight wire inductance, don't know how accurate it is but should be good enough for purposes of example. I'll use the wire you spec'd above.


    L= inductance in uH
    B= length in mm
    D= diameter in mm

    I get L = .0002*22860*(2.303*log(91440/11.11)-.08) = 41 uH

    XL=2*PI(F*inductance) = 2.58 ohms

    Add the .02 ohms for the wire resistance, and now we have 2.6 ohms. Since current^2 * impedance tells us power dissapated, we get somewhere between the equivalent of 260MW to 2340MW being dissapated by a copper wire the size of your thumb? Plus, we're talking only 2-3 circular mils per amp, typical conservative range for insulated wire would be 200-300 circular mils per amp to avoid the wire burning up. Obviously this is kinda fun to throw numbers around, but this would never work as you'd need a wire the size of an entire skyscraper to handle such monster numbers, and you're trying to conduct from two poor conductors (dirt and air).

    Sorry, your solenoid would hardly even move unless super lightweight and still move mere inches if you could get it to move without the wires vaporizing.


    Lets do the math backwards from the authors conclusions. He said there's enough power leftover for a light bulb to be lit a few months. Lets say a 100W bulb, and use a whole year. You'd need millions of good lightning captures to equal just the a single local peak demand power generator, much less a bigger generator like Hoover or Niagra Falls.

    That's not to say that lightning isn't powerful. It is. It just lasts for entirely too short a period of time and expels too much of its energy in heating air to become any kind of meaningful power source.

  10. Mar 9, 2004 #9

    Ivan Seeking

    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

  11. Mar 15, 2004 #10


    User Avatar

    Not considering problems relating constructing means for storing lightning energy captured on ground ,I must say that only 1 or 2 % of total energy of the lightning could be stored in this way.
    Some rough and simple calculations give about 50 - 500 MJ of total energy for a typical lightning discharge.And captured would be only
    0.5 -10 MJ.Not very efficient..Great deal of original energy is being lost as heat in lightning channel,dissplacement field currents in cloud and ground,far field EM radiation and so (sound waves of thunder..)
  12. May 21, 2004 #11
    Although it seems now a stale topic, with somewhat limited usefullness. I would never the less produce a simulator, that leaves an impression. The Answer to me is obvious, The lightning has a nominal pulse duration of 6uS. The corresponding resonant frequency is 1/6 us. You build a parallel tuned circuit, The L is the primary of an open circuited power transformer with the seconadary opened by means of a relay. This primary is shunted by a capacitance that equates 1/6uS=1/Sqrt(L*C). the C will be small, therefor will need to be two widely seperated sheets, of very large area to suck up the charge, as the inductance will be an open circuit at the transient, the cap a short.

    After the initial unit impulse, the tank circuit "rings", the seconadary connection is the relayedand dumped into a pulse-bumped motor-flywheel that stores the energy. It will be a cool sim, exept the visulization of the bolt is hard to simulate using a spark gap, real lightning changes direction above every 150 meters.
  13. May 22, 2004 #12
    Hydrogen production from lightning?
  14. May 22, 2004 #13
    Nitrides are the main benefit, or at least generated as a by product if lightning, and by the way, most these lighting has provided the nutrition that allow food to be grown and produced. As far as hydrogen is concerned, it is good for blimps. not much else really.
  15. May 24, 2004 #14


    User Avatar

    To me it looks like you r up to think of forcing the lighting channel to alternately "oscillate" instead of monopolary decay.
    By provoking the channel to hit larger or smaller inductance tall antenna and capacitor system for instance?That will not work in reality.Due to al least 3 reasons.You might capture small portion (a few % - best case) of the energy and that's it.
  16. May 24, 2004 #15
    I was thinking much simpler, say a simple transient circuit analysis. (1) unit pulse function into a tuned LC tank circuit, and with a delay, after the unit pulse, the power transformer secondary load relay is closed into a load, weakly coupled enough as to not totally kill the Q of the tank circuit, Or large enough to bump start a flywheel. (for a simulator of course, I am not ready to take on the gods with a kite and key, yet.)
  17. May 25, 2004 #16


    User Avatar

    There are problems I forsee in thinking "simple" in the case you want good representative for interaction between lightning
    and LC circuit of any sort.These problems concerns:
    a)Distributed geometry of the equivalent circuit representing lightning discharge (lightning channel paths are kilometers long)
    b)Impedance variation of the channel during return stroke as well as during whole event.
    c)Info good to know before the modelling in advance that there is whole class of rise times and duration times varying from shot to shot."Nominal" times of return stroke (did you reffered to the rise time 6microsec in your post?) are just nominal as word says.I don't know if that is in right ballpark or not but I suspect it isn't a fixed value.

    Even if,by some miracle,problem c gets overcome,problems a ,b include complex intraction between the struck object (Physical LC circuit) and the channel resulting in very rich frequency domain spectrum content.Impedance matching between lightning channel and the circuit wouldn' help much in capturing significant part of stroke energy eiher (the outcome is different than in circuit with concentrated and constant parameters).However,it is always better to have characteristic impedance of the circuit small (or around impedance of lightning channel) than higher due to flashower danger in the latter.
    One way or another,taking adventure in constructing means for capturing couple of % ( best case scenario) of original lightning stroke energy at one location isn't worth a try.Awerage power during one year will be too small to justify investments in the project.
  18. May 25, 2004 #17
    That would be pulsewidth 6-20 mS, risetime 50nS. Side flash would be an issue. As far as impedance matching the channel, I would not consider this an issue after the conduction path is established, It is not like a transmission line whereas you have forward and reflected waves. (maybe if it is a short circuit, ie Gamma=-1, then you can reflect the lightning stroke back to the cloud, .... not likely).
  19. May 25, 2004 #18


    User Avatar

    You have forward and reflected waves in the return stroke times.In struck object they are severe.In the channel they are attenuated due to significant disipation.
    Long lasting currents (miliseconds latter) are in order of only 100-200 amps and can be modelled I think like generated by current source.
    Feasibility of the scheme is poor.Relisation is troublesome.Ask for advice HV EE people of this forum.They should know more.
  20. May 25, 2004 #19
    TeV, you are Awwwwwssssuuuuummmm Dude. Never considered reflected waves on electrons, traveling in a plasma stream. But actually electromagnetic, effects due to the magnetic field created by such electrons provide the feedback for such phenomem.

    Actually, my "simulator", was to be a simple arc gap generator, that would be shadowed onto a larger projection area. To keep the whole thing honorable, the spark, would in fact transfer the charge to a flywheel mechanism. I thought it would be a big hit to the younger, to be able to push a button, produce a magified arc, that would actually be converted to rotational energy, maybe a wheel with a "see-through-glasses" spiral pattern. This simple demonstration is complicated enough, differential equations and all. Now I have to worry about the impedance of the source-arc for not, max, but adequate power transfer to to provide a real effect. Now add rain noise, and "thunder" effect at the pushputton, and you get a real crowd pleaser.

    Well, Ma, load up the vortex charging fuel cell in the p-truck, looks at we got to go tornado hunting again. When Ya fish, its best to go for the big ones!
  21. May 26, 2004 #20


    User Avatar

    Ok.But,in the case of a natural lightning ,simultaneous transferring of the energy in rotational (or mechanical) is even less promising as concerns efficiency than capturing it in the capacitor.Simple reason is too short time of the event.
    My only idea how to think of the problem is something like this:
    to use the system of long ( kilometers) transmission lines parallel to the ground plane.The lines should be terminated by HV capacitor banks capable of withstand megavolts.When incident EM wave reach the capacitor bank and charge it to the max potential,very fast HV switching equipment reacts and disconnects capacitor off the line.It is a horrible technical task to achive something like this in practical life.If possible ,definitively very expensive solution.Not worth a try.
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?