5

The details are hazy as I recollect this from my nipper days.Many, many years ago.

Persons were charged with stealing electricity by placing a large coil, somehow, near high voltage AC transmission lines.

Is this feasible or was it perhaps an April fool’s article that I swallowed hook line and sinker in my youthful naivety.

There is a modern reason for this question.

Due to country wide cable and transmission tower spar actual theft. We are looking at placing modules with a gyroscope and Wifi module to form a mesh that will report if a tower is being “attacked”. Of course I will still have to test if the electronic gyroscope is sensitive enough for this. But one of the problems is going to be, how to power the unit? Solar will require an extra solar panel and battery. If I could tap into the radiated energy from the actual transmission line, great. Many modern SOCs don’t require much power and can sleep most of the time.

To clarify our needs. We are looking at building a device to sense if parts of the metal tower are being removed by scrap metal thieves. If too many parts are removed from the tower. It will buckle and earth the line. This is still at the feasibility stage.

Example

kingchris
  • 722
  • 2
  • 9
  • 14
  • 2
    So you are making a device to detect if someone is sapping energy from AC transmission lines, and you wonder if you can sap energy from AC transmission lines to power it? If you weren't sure this attack is feasible, why are you building a device to detect it? – Phil Frost Aug 11 '14 at 10:44
  • 3
    No. We are building a device to sense if parts of the metal tower are being removed by scrap metal thieves. If too many parts are removed from the tower. It will buckle and earth the line. This is still at the planning stage. – kingchris Aug 11 '14 at 10:50
  • @kingchris: it would be good to add that detail to the question. – whatsisname Aug 11 '14 at 14:28
  • Have taken @whatsisname suggestion and edited question – kingchris Aug 11 '14 at 18:03
  • Would it feasible to run fibre optic cables along the spars and detect when they get broken? – Andrew Morton Aug 11 '14 at 18:50
  • @AndrewMorton: Yes. But it would take too long to wire say 400km of an existing line with fiber. Google a pylon picture. Normally spars/beams from the four legs are removed. The power company would prefer a single self contained unit that has to be attached somewhere safe to each pylon. As I said elsewhere they might only "arm" pylons near humans. So your idea could work if the number of pylons is low. – kingchris Aug 11 '14 at 19:03
  • http://ieeexplore.ieee.org/xpl/login.jsp?tp=&arnumber=4773888&url=http%3A%2F%2Fieeexplore.ieee.org%2Fxpls%2Fabs_all.jsp%3Farnumber%3D4773888 <-- this may prove useful later. If you mount gyros to the tower its self you will need to contend with some radiated form of these vibrations. – Nick Aug 11 '14 at 20:09
  • Any reason you want to go gyro rather than an accelerometer? I would think a thief would be doing more vibrating on it rather than twisting of the structure. – horta Aug 11 '14 at 20:25
  • 1
    @kingchris Hopefully you have patented key aspects of your ideas. The really really down-under NZ patent system may be your friend - others may or may not work like this. For $NZ50 you can file a provisional patent with 12 month lifetime. You can simply make a file with all your original material in it (and any other material) and submit it. Within one year you can write a 'real' patent - any ideas concepts whatever that can be shown to be wholly based on the provisional acquire its priority date. I am not even remotely close to being a patent lawyer - if this sounds useful ask a real one. – Russell McMahon Aug 11 '14 at 22:50
  • @horta : Eventually it will come down to cost, sensitivity and range of values. We will experiment. Biggest problem is to map normal range of pylon movement so the module knows when to "Phone Home" to warn that it might be under deconstruction (:-) – kingchris Aug 12 '14 at 04:20
  • 1
    @RussellMcMahon: Lawyers and patents. Difficult. Cost an arm leg and one kidney. If someone violates your patent you have to give up the other kidney to take them to court. I am sure that there are probably existing seismic monitoring units with Wifi that can be re-tooled, recoded. Ideas are a dozen a dime. Getting them to market working. That's the 99.999999999% effort. Then the patent trolls come out of the wood work anyway. But good to know the really really down-under has a useful patent system. – kingchris Aug 12 '14 at 04:27
  • @kingchris For interest only - any progress on this? – Russell McMahon Jul 14 '15 at 20:31
  • A commercial product was developed called CapTap that was connected to the line and used a capacitive potential divider to reach much lower voltages, I don't remember how high a voltage it was rated for but it could provide quite some current, perhaps there are ideas to be had there. Did not find mention on the net, perhaps 20 years ago. – KalleMP Jul 14 '15 at 21:11
  • @RussellMcMahon No movement as they say. The idea of having self powered IoT devices attached to a pylon was accepted but the logistics of humans placing these devices was the deal breaker. So back to the drawing board. We now need a device build as a spider that will climb the pylon and self position. And they talk about feature creep ! – kingchris Jul 16 '15 at 04:58
  • How about welding things together? Then people could not unbolt parts, they would have to cut them. –  May 12 '16 at 20:28
  • @nocomprende that would be a good idea but the problem is the existing towers that were bolted together. – kingchris May 13 '16 at 05:28
  • 1
    So go around with a portable welding rig and lay some weld lines? Probably a lower tech solution than installing a complex monitoring system, which would *still* require going after the perpetrators. It is usually better to make things impossible than to come up with a bigger stick to beat people. Sorry to torpedo your concept, but I think this is a better idea. –  May 13 '16 at 13:24
  • @kingchris Another 7 years on :-) - did anything eventuate. A feature creeping spider sounded like fun. – Russell McMahon Jun 13 '22 at 11:36
  • @RussellMcMahon. It was a think tank type problem that funding never materialized for. So a POC stage didn't happened. I do enjoy the slightly crazy ideas and hunt for solutions, but one really needs that fat grant or billionaire who will write you the mythical blank cheque. – kingchris Jun 16 '22 at 06:04

5 Answers5

11

“Persons were charged with stealing electricity by placing a large coil, somehow, near high voltage AC transmission lines.”

Is this feasible or was it perhaps an April fool’s article that I swallowed hook line and sinker in my youthful naivety.

Entirely feasible.
Farmers were occasionally charged with power theft in this country (New Zealand) in the past. I haven't heard of a case in a decade or few - maybe they are getting cleverer at it :-).

This is the same principle as used for "IPT" / "Inductive Power Transfer as seen in phone chargers, industrial monorail powering, electric vehicle charging and much more.

I started to say that if the pickup coil was symmetrical with respect to two phases that were perfectly balanced that you'd get zero pickup, and then suddenly realised that I've always done IPT with essentially a single phase, and that with a 3 phase system with 120 degrees phase separation you should get the advantage of the full load current even if the two phases were fully balanced.

You are essentially getting fields produced by the current, not the voltage, and the voltage is essentially irrelevant as long as you observe the normal conventions that apply to any other dealing with xxx kV.


Energy Harvesting from Electromagnetic Energy Radiating from AC Power Lines - FAR more energy can be obtained than they achieve.

Worked example - I suspect some of the conclusions are suspect A Solution to the RWP for Exam 1 - Stealing Power

Low technical content - high relevance

Directly relevant but low technical value Electromagnetic Harvesters: Free Lunch or Theft!

Several related stack exchange questions with variably useful content.


Online vehicle transfer - I do not have access to this paper but it is probably at least relevant as it will have examples of dual linear conductors and a pickup coil.

Mythbusters getting it wrong


Related:

Industrial monorail

Maximising transfer


Capacitive - but impressive:

enter image description here

Russell McMahon
  • 147,325
  • 18
  • 210
  • 386
  • 2
    Do you not have have enough work to do, really really down under, (NZ) near the south pole. This is an excellent answer. Must of taken you some time and effort. Thanks a lot. I am not with the power company but hope to sell them a gazillion units. The problem is that they may only buy for pylons near human habitats as that is where the thieves may reside. Again excellent answer – kingchris Aug 11 '14 at 18:13
  • 1
    If you're hoping to sell the power companies a product you've got a LOT of R&D to do. Starter question: What happens during a lightning storm? – John U Aug 12 '14 at 13:35
5

In my final year (1965) of my studies for my electrical engineering degree, I undertook a project on behalf of the then New Zealand Electricity Department to investigate the feasibility of extracting power from a 220 KV three phase transmission line in order to power a radio repeater located in the Southern Alps of New Zealand's South Island.

The theoretical analysis was performed by Mr Jack Woodward, then Reader in Electrical Engineering and acting Head of Department of Electrical Engineering at Canterbury University, Christchurch. The setup consisted of a conductor located about 10 feet below the outside HV conductor and 10 feet away from the vertical plane containing that outer HV conductor.

The manual computations for just one possible location of the pickup conductor took approximately four hours, accordingly, a number of programs were written in Fortran IV for the university's IBM 1620 computer to perform the calculations at each point in a grid in the plane perpendicular to the HV conductors.

The principal results were that extraction of sufficient power to run a radio repeater was feasible, however the pickup wire needed to be as close to the HV conductor as regulations and safety would allow (one had to allow for additional sagging due to the weight of snow on the conductors). At this point, the electrical potential was high (about 20 kV) hence an instrument potential transformer was used to deliver about 230 volts to the load. Some inductive tuning out of capacitive reactance was needed to improve voltage regulation. Following this computer based investigation, NZED engineers and technicians carried out field tests which confirmed the conclusions drawn by the study.

One interesting side note is that the calculations involved inverting a 4 x 4 matrix. The 1620, as I recall, took about 20 seconds to perform this task. My current desktop computer will invert higher order matrices in a fraction of a second.

The computer programs I wrote back in 1965 no longer exist, but I am presently trying to re-create them in order to compare execution times.

pipe
  • 13,748
  • 5
  • 42
  • 72
Kevington
  • 51
  • 1
  • 1
  • Well keep us posted on the calculations and perhaps show the theory behind the maths. Thanks – kingchris Oct 11 '16 at 05:51
  • For the theory, i refer you to material such as http://services.eng.uts.edu.au/pmcl/pct/Downloads/Lecture05.pdf. – Kevington Oct 11 '16 at 08:32
  • Modern desktop CPUs can invert a 4x4 double precision matrix in something like 100ns and that’s very conservative for just one matrix ready in cache. If you got many of them, the pipelined rate can be higher. That’s just one core. On mid-range GPUs, the sustained inversion rates of half or single precision inversions will be main memory-bound. The throughput for matrices in VRAM would be on the order of 1ns/matrix. – Kuba hasn't forgotten Monica Apr 27 '23 at 12:22
4

Electrostatically Harvesting Power from High Voltage Power Lines. Note from author: Please do not try this experiment without a through knowledge of electricity! Please understand that it is illegal to take power from power lines. This is an experiment only.

If you run a single wire along a high voltage power line, (= or> 340KVAC), you will harvest a fair amount of voltage at micro amp levels. WARNING: KEEP AT LEAST 30 FEET FROM THE HIGH VOLTAGE LINE!! DO NOT DO THIS EXPERIMENT ON WINDY DAYS!! The longer the parallel wire, the higher the voltage and amperage you can collect. Run your (antenna wire) parallel to the power line in the trees adjacent to one outer phase of the power line. The higher the wire, (antenna), the better. Use small gauge bare wire hung from fishing line for insulation and support. Run at least 200 feet of bare antenna wire. 400 feet equals twice the power collected. Use a fishing pole and weight to put up the wires in the closest branches of the trees next to the HV line.

WARNING, DO NOT TOUCH THE ANTENNA WIRE! YOU WILL GET SHOCKED! I recommend two pairs of rubber gloves and good boots for this. Ground the Antenna wire while you work until the transformer (s) are installed and you are ready for power. Provide a single or several 5 foot copper-plated ground rods at your termination point for the transformer high voltage return to ground. More grounds are better. The ground completes the high voltage antenna circuit.

The trick is to present a minimum load to the antenna wire and to step down the electrostatically induced voltage. We need a very high voltage transformer to do this. A ham radio plate transformer rated at 10,000 volts works good. Neon sign transformers work well. You can use several used oil burner transformers with the secondary in series. They are 10 KV each so you can make a 50 KV secondary with five of them. Much better! Remember, the higher the transformer ratio from the high side to the low side of the transformer the better match you will have to the source (the power line antenna). We then connect the high voltage transformer secondary (HV) winding to the antenna and the ground. Separate transformers MUST have the 120 volt leads in parallel and IN-PHASE. You have created a capacitive-coupled step-down transformer. Place a 4 watt, 120 volt light across the 120 volt terminals of the transformer. If the light pops and you measure a higher voltage, you may add another single transformer to get to the voltage you desire from the HV transformer (s). You need to measure the AC voltage out of the HV transformer primary and find a 120 volt transformer for that voltage. Example: 600 volts out requires a 600 volt secondary and a 120 volt primary. Place this step-down transformer after the HV transformer. You may also ground the far end of the antenna wire to create electromagnetic coupling to the power line. No other circuit changes are required for this. Bring along some different wattage/voltage lamps, a car battery charger and small battery and some small AC motors to see the power you have collected. Have fun! Author: I am purposely not saying I ever did this. Work safely and remove the antenna when done so it will not blow into the high voltage power line.

Nick Alexeev
  • 37,739
  • 17
  • 97
  • 230
  • Might be fun to try this. I could also read up on the Tesla papers where all sorts of other interesting energy harvesting ways are listed apparently. – kingchris Oct 11 '16 at 05:54
2

Yes, it's feasible (although entirely illegal - unless you work for the power company...). It's just an air-cored transformer. Get a coil into the magnetic field radiated by the transmission line and off you go.

That said, I'd do some sums and experimentation before hooking any electronics up to the output; you don't want to feed 10kV into your 3V embedded system...

markt
  • 4,936
  • 2
  • 15
  • 13
  • You are correct. Turning 10k AC into +/- 3.3v DC will be an interesting problem. – kingchris Aug 11 '14 at 10:41
  • Should be manageable enough. Start by finding out how far away you will place your coil (safety first!!!). Then calculate (and verify by measurement) how much voltage you get per turn of coil of a given diameter at that distance. As long as you aim for less than 32Vac (12Vac or so would be ideal) you should be fine. You might want to have a chat with the transmission engineers too, they can probably help with the calcs and should be able to tell you how much leakage will trip their power theft alarms (thefts are detectable and locatable due to the change in line impedance). – markt Aug 11 '14 at 10:52
0

Here is another thought: You can make your device a live tank device that clamps to the line its self and operates with a current transformer. Your whole device would operate at line potential. You would not need to include insulators for mounting (which are expensive and not trivial to design/produce). From there you could do some sort of motion sensing or video looking down at the ground around the base of the tower.

You would want to filter our line vibrations to keep from causing a bunch of false alarms. There are some IEEE articles on the types of frequencies power lines vibrate at due to wind loads.

From there you can relay the data off via radio, fiber, or PLC Modem.

Doing it this way allows you to design a greedy system that never needs to operate in a sleep or low power mode. You don't need a battery (if we assume these transmission lines are always carrying a few amps of current), and you don't have to worry about any capacitor charging schemes. You just need a coil, a rectifier, and a DC/DC converter - then you are in business.

This type of design is also inherently immune to lightning because there is no ground path. You just need to design to protect your system to shunt current spikes that could be observed due to switching, faults, and lightning (if the current passes through the primary of your CT en-route to an arrester.

The downside: Maintenance and installation are tougher. A lineman will have to bond to the line to install it. Otherwise it needs to be hook stick installable.

Regardless of how you design your system: Look into the SCADA protocols to the utilities you market to already use. If you can use an IEC61850 or DNP3 library and the utility can directly integrate your tool with their SCADA system it will be more marketable. Your customers would likely be more receptive of something that ties into the system their dispatch center already uses, rather than some new piece of software they have to run on some machine in their central office.

Nick
  • 335
  • 1
  • 12
  • Not sure about this. We want to Wifi gyro movement above a specific threshold to the next unit and then onto the next eventually to a server node. Having the unit on the cable means the 10k to 30k alternating field may screw up the wifi signal. We are not yet sure if a standard Wifi module will work close enough to a power line that we can tap into its alternating signal for power. Much testing to be done. The SCADA idea is an excellent one. Thanks. – kingchris Aug 11 '14 at 18:46
  • Wifi works just fine on overhead 30kV distirbution class breakers powered using the same method. Not sure if the game somehow changes at a few hundred kV. The bigger issue I think you would face is range when using wifi a hundred feet up in the air if you are trying to talk to it from the ground. There will always be a certain amount of vibration around the transmission tower due to wind - regardless of whether you do some image processing to detect certain types of potion, use gyros, accelerometers, or whatever, you will need to filter / discard the wind induced vibrations. – Nick Aug 11 '14 at 18:54
  • Case and point: http://www.hubbellpowersystems.com/catalogs/switching/10E.pdf#page=7 , But like I said - I think range will hurt you w/ because ground to tower top height can be quite high. – Nick Aug 11 '14 at 19:02
  • Thank for the info. I feel that we will have to build a unit or two then go off to a local line and start testing with the power companies blessing. Range may well be a problem. – kingchris Aug 11 '14 at 19:06
  • But - like I said - you can do this in another way. You can install this device 'looking' down at the tower bottom (because your metal thief is likely not going to climb up the tower too much, and even if he did, you'd see him). Then use a video / PIR sensor of some sort and try to see if you detect a certain magnitude of motion around the base of the tower and exfil a photo or video whenever you think someone is moving about under the tower. – Nick Aug 11 '14 at 19:08
  • Good luck with getting their blessing to install stuff on their transmission lines :) Most power companies are pretty funny about their transmission infrastructure - at least around here. Why not see if you can test it in a substation instead? You're much closer to the ground then - it might be a little easier to sell. – Nick Aug 11 '14 at 19:09
  • Sure. Its all at the idea stage for now. – kingchris Aug 11 '14 at 19:11
  • @Nick I suspect the local wildlife would exclude the use of video/PIR sensor methods - too many false alarms. – Andrew Morton Aug 11 '14 at 19:29
  • @Andrew true. But perhaps some low cost chicken wire fences could mostly solve that problem. But regardless of wildlife, if you used video you'd have to filter out wind blown shrubs, sunrise/sunset, as well as the fact that your platform its self is moving w/ line vibrations. I never said it would be easy! But in was mentioned earlier that gyros presented some difficulties. So I was throwing up an entirely alternative idea to spur thought. – Nick Aug 11 '14 at 19:41
  • @Nick My search for "rhino vs chicken wire fence" didn't turn up any matches(!). Another thought was a microphone attached to the structure to sense cutting/unbolting, but, as you point out, wind effects might make that a useless idea. – Andrew Morton Aug 11 '14 at 20:18
  • The key is motivation of the animal. I have family who own a cattle ranch. Chicken wire fences control the cattle nicely most of the time. But they don't work as well when you separate a calf from its mother for the purposes of 'brain surgery'. At that point the fence is no match for the very motivated mother. But I digress :) – Nick Aug 11 '14 at 20:27
  • I don't see how the current passing through the line will induce any current in the coil if the coil is at line potential. – Milind R Nov 08 '19 at 19:11
  • The secondary is ungrounded and is effectively floating. But in this case the primary a distribution or transmission line that’s energized with 10s of kV. So everything is energized at close to that level. However a potential will form across the secondary proportional to the current. – Nick Nov 09 '19 at 23:42