> Thicker wires will heat up less for a given amount of current, increasing the carrying capacity of the circuit. One solution to increasing the effective thickness of a conductor is to “bundle” several conductors a few inches apart from one another, allowing for a larger increase in current for less cost than a conductor that is simply double the size.
In power engineering at university we were taught that the separation of the wires within a conductor, to increase the effective diameter, is primarily about reducing the self-inductance rather than dissipating heat. From memory, it's along the lines that the separation effectively increases the diameter of the conductor, which in turn reduces the intensity of the magnetic field, reduces the self-inductance and allows more power transfer for a given voltage and current.
The article got it right in saying spacing the wires gives an increase in current for a given cross-section of conductor, but the main reason is reduced inductance rather than the reduction in resistance due to it running cooler. There will be some reduction in resistive losses due to cooler conductors, but the main gain is the reduction in reactive power, which otherwise causes current flow with no power transfer.
(I'd edit my post, but its long since too late for that... mea culpa)
There's a couple of things being mixed up here. Stranded wire versus bar is a mechanical issue. But what you were talking about is the separators you sometimes see to run several separate cables alongside each other in the same phase. In this case, there are a couple of things being overcome. For starters, the skin depth of Al and Cu is only about 10mm at 50-60 Hz. However, the form of the limit is not the reactive power loss, its the resistive power loss of the conductor. Eventually, making it thicker just stops decreasing its resistance at that frequency.
The article hinted at this in the discussion on cable construction, when the author was describing wraps of aluminum conductor around a steel core. Now, imagine an A/B comparison, where the A side is a very large diameter cylinder of Al strands around a large steel core. The B side is three smaller such cylinders separated by air, with the same total circumference of Al. The B side is much lighter overall for the same current-carrying capacity. You might like to hollow out the A side, except that you need to endure the compressive stress of the outer sheath of Al conductors bearing down on the structural core somehow. Enter the periodically spaced triangular separators that actually are present on the B side.
This is technically true, but a smallish effect at the scales its employed. The biggest reason for stranding transmission conductors is that its easier to bend them.
Long spools of wire are much easier to transport and assemble than bars.
I was always taught that one of the main reason to use multiple conductors was due to reduce the losses caused by the corona effect of ionizing the air.
As another commenter says, bending might also be a factor.
I read this wonderful article a few years ago and have been looking for it ever since.
I recommend it highly to every engineer who builds software, hardware, or whatever. A great story of what can go wrong (very wrong) and how even the best attempts to fix it may not work and may in some cases make things worse.
It's superbly written and very educational. The "Age of Electrical Outlets Having a Little Face and Different Sized Eyes," indeed. :-)
Here in California, the State government has maps of all of the transmission infrastructure, from the generating stations to the utility substations, and the transmission lines in between.
See also https://openinframap.org/ which pulls from OpenStreetMap. (I'm one of the contributors to the mapping of the Bay Area's electrical infrastructure.)
What you're seeing on that site isn't actually OpenStreetMap. It's data points from OpenStreetMap that are overlayed on top of Mapbox [1], with the vector tiles served up by Tegola.
Neat. Lots on electrical, though not much on gas pipelines (natural and gasoline). There are some strange ones too, such as fertilizer, that are not mapped.
If you like this type of data, I've been working on an ArcGIS app[1] that incorporates the USDHS transmission infrastructure maps[2] from the western United States (WECC region) with NOAA weather data, US Forest Service wildfire data, and 15-minute market prices from the California ISO's OASIS system.
There's a lot of work to be done and I've used this project to teach myself Python with the ArcGIS Python API, Pandas, GeoPandas, etc.
I'd love to hear feedback from anybody that has something to share.
Strangely, at least in the case of vital infrastructure like this, making the information public also poses a liability. I found it strange as I figure bad actors could surely find this very useful.
>Moving into the future, it’s hard to say how much more modern the power grid can get since the underlying principles are so simple: three phases per circuit and structures large enough to keep them from sagging into something that could cause a fault. There is a lot of talk of the smart grid, but the solution to most of the issues with the power system is often simply to build more circuits as the demand for electricity rises. It’s a difficult problem to engineer ourselves out of, especially with the increasing age of the power grid itself, and at some point is simply becomes a numbers game of how many watts can be moved from place to place.
I wonder if this will be a tough sell to the generation that grew up playing Sim City 2000, which had this kind of system going haywire as a disaster scenario.
There is a book from the mid 1960s (I think the last revision, a few older ones) that is still basically current (hehe) by Westinghouse called "Electrical Transmission and Distribution Reference Book"
Thanks for the recommendation. For a more modern take, the chapter on "The Power Grid" in Brian Hayes' "Infrastructure" is tremendous as well: http://industrial-landscape.com
AC power has a voltage that (in North America) oscillates high and low at a rate of 60 cycles per second.
Three phase AC power sends power along three wires. The timing of the oscillations on each wire is shifted by one-third of a cycle relative to the other wires, so they all reach peak voltage at slightly different times. We call this a "phase shift" between the voltages, or that the voltages are "out of phase".
If these wires were connected to a 3-phase AC motor, the motor would spin because the current is drawn in succession along one coil and then the next, leading the rotor to follow. The motor coils are also called "phases".
Each wire is sometimes called a "phase" as shorthand.
Back in highschool shop class I was tasked with building a three phase motor demonstration. The idea was to visually demonstrate the rotating field of the the three phases in the motor windings for open house. I took an old 1 horsepower (0.75kW) motor, pulled the rotor, and sat a few large-ish ball bearings in the stator housing. When plugged in to three phase the bearings rapidly spin around the inside of the stator housing.
I had to do a few more mods such as clear plexi end covers as well as add resistors in line with each phase to drop the current. Without the rotor the inductance is much lower pulling way more current than the nameplate rating (otherwise magic smoke). Then wired it to a contactor and button to run it.
In a very small nutshell: to get continuous power but still use AC it helps if you offset the AC on several conductors. The lowest number of conductors with which this works well is three, and 360 degrees rotation offset in three equal bits is 120 degrees, hence three phase power, where each 'phase' is offset by 120 degrees from the other two. The current and voltage between the conductors are said to be 'out of phase' by 120 degrees.
Well this was an educational read, thanks for sharing. I particularly enjoyed finding out about surface transmission characteristics of alternating current and ACSR bundles. More on HVDC would be good.
It was uncomfortable too: I feel about power lines the way some people feel about snakes or heights. I find being near or under them extremely disconcerting. I even have an infrequently recurring nightmare where I find myself having to crawl near one of the thick cables. They make my skin crawl.
The data center I work in has 415 VAC power coming in to each rack, with three phases and a neutral. Equipment gets a phase and a neutral, for 240 VAC (though not in a configuration you’d normally expect).
It’s important to remember that everything is contained in equipment made by companies that I trust (Rittal and Eaton and APC). Also, everything is grounded (the doors to the racks, the racks to the cable ladder, and the cable ladder to a special ground), which is in addition to the equipment ground. Finally, I know (and work with) the people who build and maintain the infrastructure.
As for the thick cable, if you’re up to it, here’s a video showing a process for joining the cables together: https://youtu.be/EvWx-VKVvmo
The thing to note, is the layers of copper ground that are wrapped around each conductor line, as well as the bundle itself.
Thanks. Had to do some reading to understand that, made more sense once I learned the phase-neutral voltage was the supply voltage / sqrt(3) not just divided by 3. Then I started reading about wye configurations and now it's time to stop.
There's a marvelous video somewhere on YouTube showing someone splicing a _live_ conductor while dangling from a helicopter. Can't find it now. Yours still gave me the willies though! Horrid big deadly snakey things. Ugh.
Power transmission in China has passed the 1 megavolt level. 12 gigawatts. China's best energy sources are in the northwest, and the biggest loads are in the southeast. Hence the demand for very long transmission lines.
Agreed. There's two things about HVDC worth mentioning:
It is High-Voltage because high voltage is efficient for transmission. On a first-order approximation, resistive losses are proportional to current, not power. Since P=VA, the resistive losses decrease for a given power as more of the power shifts from the A to the V.
Second, a neat property of DC transmission is it allows phase-decoupling between grids. For you to get a clean 120V, 60Hz power line at your outlet, all the transmission and distribution infrastructure needs to be tightly balanced. When that 60Hz starts to deviate by like 0.01%, alarm bells start going off at the utility. Now imagine keeping that AC stuff sync'd on a state-wide transmission line. If you make the line DC instead, you can transmit power willy-nilly across the length of the line and just make sure it syncs up at the ends when you change it back to AC.
We haven't traditionally used DC because it has higher transmission losses than AC until you get up to the ultra-high voltages (that's famously why Edison's AC system won out to Tesla's DC in NYC back in the day). My understanding is it's just been really hard to get up to the ultra-high voltages needed to make it work. 12V DC transmission barely works at even house-scale because of the resistive losses, but when you get up to 1MV like they're building in China, it starts to get attractive, especially when linking different sub-grids half a continent apart.
Volt for volt, doesn't AC have higher transmission losses than DC because of the skin effect? I was under the impression AC was what we chose for the grid because transformers make it cheap and easy to step up and step down voltage.
"Early DC power stations had a dynamo for every light bulb. Source unknown." No way. Edison's first Pearl St station, which is very well documented, did not work that way. Edison's Menlo Park NJ demo did not work that way. Some really early arc lamp plant, maybe.
There's a problem with paralleling DC generators. Edison hit that and solved it. Scaling problems, the early years.
I've worked in big piles of 3-phase power cables (though not quite as fat as in those other videos!). It's not bad. It's not any different than working with radiation, or at height, or around table saws, or crossing the street. You learn how it works, and how it can go wrong and hurt you. You practice good work habits, and learn to trust your equipment and training and coworkers. You recognize that it would take 2 or 3 things all going wrong at once for it to really fail, so you report any 1 issue right away.
The transmission towers might be easy to design for flat lands and plateaus but not so much for hilly areas. I am a always fascinated by transmission towers in the mountains, wether I’m driving on the highway and see them far away or I’m hiking and run into one. I’m amazed by how they get such big structures up and running in those areas.
Edit: I just think the author kind of rubbed them off is all.
In 2015 the Center for Land Use Interpretation did a bus tour of high-voltage infrastructure in Southern California. The write-up has some interesting photos:
When the power is generated, it is usually done so by a 3 phase generator that is separated by 120 degrees of rotation. All 3 phases inside the generator are bonded together on one end. The other end are the 3 current carrying conductors leaving the generator. The bonding of the phases on each side of the conductors is what creates the closed circuit.
In power engineering at university we were taught that the separation of the wires within a conductor, to increase the effective diameter, is primarily about reducing the self-inductance rather than dissipating heat. From memory, it's along the lines that the separation effectively increases the diameter of the conductor, which in turn reduces the intensity of the magnetic field, reduces the self-inductance and allows more power transfer for a given voltage and current.
The article got it right in saying spacing the wires gives an increase in current for a given cross-section of conductor, but the main reason is reduced inductance rather than the reduction in resistance due to it running cooler. There will be some reduction in resistive losses due to cooler conductors, but the main gain is the reduction in reactive power, which otherwise causes current flow with no power transfer.