BEFORE you jump to answer or flag as duplicate read please read this first! I always assumed like yourself that AC would be more efficient and in part that is what Tesla and Edison fought over to reconcile and AC won and the rest is history....enter. "Scientific American" page 20 June 2017. Article titled " Electric Renaissance" by Annie Sneed. says

" Technology for power transmission advanced in the 1970's, allowing direct current to return as a viable option-and for lines more than 300 to 500 miles long, DC out competes AC.After a certain distance , AC systems become more costly to build than DC and have larger power losses along the line because of issues such as higher resistance. " Using DC lines is a much better solution for moving power from big, remote wind or solar farms",says Gregory Reed, director of the University of Pittsburgh's Center for Energy and Energy GRID Institute...."

I always thought DC sources were converted to alternating current for efficiency sake not to just accommodate the AC infrastructure that is all ready in place. Obviously you have to step down the AC before it is allowed to get in side the house but isn't the reason DC was never used to begin with was because of the resistance in DC power transmission and that the lines would have to be as thick as a telephone pole not to mention you need a power source at every corner. Can some one clear my obvious confusion. Thank you again!

  • 3
    \$\begingroup\$ Paragraphs are definitely more efficient than walls of text! The key issue is that high voltage is more efficient, regardless of whether it's AC or DC, and until recently it was much easier to do voltage conversion with AC. \$\endgroup\$
    – pjc50
    May 27, 2017 at 19:43
  • 3
    \$\begingroup\$ High voltage AC was originally used for long transmission lines because the technology to efficiently convert from low to high DC and back again did not exist. Now that it does, DC is used. for long transmission lines because it is more efficient. \$\endgroup\$
    – Barry
    May 27, 2017 at 19:44
  • 2
    \$\begingroup\$ "The HVDC Inter-Island link is a 610 km (380 mi) long, 1200 MW bipolar high-voltage direct current (HVDC) transmission system... commissioned in April 1965. The link originally was a bipolar 600 MW link with mercury arc valves" - en.wikipedia.org/wiki/HVDC_Inter-Island \$\endgroup\$ May 28, 2017 at 3:46

6 Answers 6


HVDC is more efficient over such distance. Here are the reasons:

There is no inductive/capacitive reactance in the case of DC whereas in case of AC both capacitive/inductive reatances exist. Due to the absence of Inductance, the voltage drop in HVDC is very small as compared to AC (Provided that all other conditions are constant). Due to this reasons, the HVDC provides an edge over the voltage regulation of power system.

HVDC is free from dielectric losses. Also, there is no skin effect of conductor and whole conductors is utilised for power transmission.

In terms of cost of power system DC is more efficient and is least expensive. Firstly we only require two conductors instead of three. But the only problem we suffer is a generation of power at high voltages because we need HV for transmission. DC can neither be stepped up directly and neither can it be generated at very high voltages. So at present, the only solution is to use Solid state electronics and convert HVAC to HVDC for transmission purposes, then again convert it back to AC and step down for transmission purposes.

  • \$\begingroup\$ if you can't generate it at high voltage then you still need to depend on AC for the high voltage and use DC just for the transmission part but keep the AC infrastructure for stepping down after the DC to AC conversion. So Edison had half the answer and Tesla had the other half. \$\endgroup\$
    – Sedumjoy
    May 28, 2017 at 4:25
  • 1
    \$\begingroup\$ @Sedmumjoy Actually, operating HVDC lines do not rely on the AC infrastructure in anyway. In fact, it's a nuisance. They take AC, rectify it to DC, use a high voltage IGBT stack to switch it at ~10-20kHz through a high frequency transformer, then rectify it again. If the entire grid were on DC, it would save the initial rectification step. Also, Edison did not have half the answer, nor did Tesla. Only recently has technology enabled HVDC. AC was the only grid transmission option for well over 100 years. Old radios had to have batteries to provide DC, there were no practical rectifiers. \$\endgroup\$
    – metacollin
    May 28, 2017 at 11:29
  • 1
    \$\begingroup\$ There is no absence of inductance. Inductance is present and unchanged regardless of what kind of current is flowing. \$\endgroup\$
    – metacollin
    May 28, 2017 at 13:05
  • 1
    \$\begingroup\$ Sorry, but you have completely overlooked the conversion losses. \$\endgroup\$
    – winny
    May 28, 2017 at 19:43
  • 1
    \$\begingroup\$ "Firstly we only require two conductors instead of three." - 2-wire DC only carries one third the current of 3-wire (3-phase) AC, so that is not a valid reason. If it were, DC would be an attractive option at short distances, too. \$\endgroup\$
    – marcelm
    May 28, 2017 at 22:18

1. Tesla ≠ God, Edison ≠ the Devil

(To paraphrase an excellent Forbes article title.)

The War of the Currents was between Thomas Edison and George Westinghouse. Tesla was involved only so much in that Westinghouse licensed his patents on AC power transmission. It's worth adding that Tesla actually worked for Edison, and used Michael Faraday's prior development of AC current and reworked Edison's own DC power generation and transmission technology to use AC instead.

And he did so at Edison's request.

Edison did not have much faith in DC power, and asked Tesla to try to redesign it into something better. Not something that used AC specifically, just something. Something that was better. Which, obviously, Tesla did. However, Edison thought the idea of using AC for power transmission was too outlandish to be practical and chose not to pursue it. This was...unwise of Edison to be sure.

This would be the last time Tesla and Edison had any interaction or confrontation with each other over DC vs. AC power transmission.

George Westinghouse, on the other hand, truly believed in AC power, and while Tesla refined and patented many AC-related technologies, it is Westinghouse that commercialized the technologies Tesla invented into something that could be practical for power transmission. It was Westinghouse that pushed for AC, funded it, built it up, paid to have those shows demonstrating the danger of DC, all of it. Tesla continued to help with the development of AC to a limited degree, but he did so because Westinghouse paid him to, he was not particularly passionate or concerned about making AC the global standard. Remember - Tesla thought power transmission using conductors was itself fundamentally flawed, and was heavily focused on developing wireless power transmission. Unfortunately, this ultimately proved to be wholly impractical, but I digress.

To highlight this, Tesla spent the final years of his life in The New Yorker Hotel, one of the last bastions of DC power. It operated a large DC power plant until the 1960s, long after Tesla's death, and was, during Tesla's last years, one of the largest (and last remaining) locations of DC power generation in the world. Choosing this location as his home and place of work until he ultimately died are not the actions of someone who was particularly concerned about AC vs. DC power. Tesla only wanted to prove his ideas could work.

The whole idea of Edison vs. Tesla is mostly fiction, a fun story borne out of the internet, but sadly, not one which reflects history or reality.

See here for a historically accurate, non-fiction version of the War of the Currents.

2. HVDC and HVAC

Now, to finally answer your question, HVDC is substantially more efficient than HVAC for power transmission.

This is why new transmission runs are exclusively HVDC today. It simply makes no economic sense to build HVAC transmission lines anymore. HVAC still dominates of course, but because this sort of infrastructure has very long service lifetimes. Anything new that is built is always HVDC.

Note, the key word here is transmission level runs. These are the extra-high voltage long distance regional runs. AC is still the best choice for things at the power distribution level, but that might change (or might not).

So how much more efficient is HVDC than HVAC? Sit down - HVDC is between 30 to 40% more efficient at power transmission than HVAC. So it isn't just a little better - it's a lot better.

Now, lets talk about why:

  1. Reduced Corona losses


    Since HVDC is of static polarity, it lets charge carriers arrange themselves such that the electric field between conductors is minimized, resulting in roughly half of the losses due to corona currents compared to HVAC.

  2. No Skin effect

    enter image description here

    See how those high voltage lines are actually made up of bundles of 3 conductors? That's because of the skin effect. Alternating current tends to concentrate the actual charge carriers (electrons) of a current in the outer-most regions of a conductor. This effect is dependent on the frequency and conducting material, but it effectively makes a given conductor behave like a somewhat smaller conductor when AC is put through it. For a 1000 square mm cross section of copper, 60Hz AC will see the cable as roughly 23% more resistive compared to DC current through the same cable.

  3. No Reactive Power

    enter image description here

    Reactive power is the imaginary component of complex power, a property unique to AC current. Reactive power, unlike active power, does not represent actual transmission or consumption of energy at a given rate, but rather the rate of energy storage. Reactive power represents power that is doing nothing, it is just power that is feeding parasitic energy storage mechanisms, specifically capacitance (which stores energy in an electric field) and inductance (which stores energy in a magnetic field). This reactive power averages out to zero, as the stored energy is eventually released back (or, is negative in magnitude). The problem is it doesn't actually do anything, it is not usable by a load, it is not really transmitting any power to to a distant consumer. But, the actual currents involved are very real, so with AC, there is a component of the current that is flowing and suffering resistance losses and all of that just like any other current would, but it is doing so uselessly. It's just wiggling around without doing anything, and those wiggles translate into extra losses that HVDC doesn't have.

  4. No Dielectric Losses

    enter image description here

    Dielectrics, also known as insulators, are made up of lots of little electric dipoles (hence, 'dielectric'). As the polarity of AC changes back and forth, so too does the electric field. This makes the dipoles in dielectrics want to flip around to a polarity favorable for a given electric field. Flipping these little dipoles isn't free however, it takes energy. This means that dielectrics in bulk have the effect of dissipating a small amount of electromagnetic energy when they undergo polarization by an external electric field.

    In DC, this happens once, when the current first begins flowing. After that, the polarity doesn't change (unless something has gone terribly, terribly wrong). For AC, it changes at whatever the frequency is. For grid transmission, that is usually 50 or 60Hz. You might think, "but the power lines are bare metal, there isn't any insulation!". Actually, they are all wrapped in insulation, it just happens to be air. It's not my favorite insulation material, but it sure is cheap! This effect, while very small at 50Hz or 60Hz, when multiplied to long distances, high electric field strengths, and just shear scale, it all adds up into another meaningful source of loss. One that HVDC doesn't have.

  5. Phase? Frequency? What's that? enter image description here

    HVDC is not a sine wave, or any wave for that matter, and does not have phase. It also doesn't have frequency.

    The whole point of a power grid is to serve as an interconnect not only between power station and power consumer, but between multiple sources of power generation simultaneously. This has countless advantages, from redundancy to load sharing and balancing.

    You can't just connect two grids up, wrap the cables with some electrical tape, and call it day. The grids must not just be AC, but AC of the same frequency and in phase. Worse, there are areas of the world where adjacent grids that need to be connected aren't the same frequency at all, none-the-less in phase. This all adds up to make the interconnection of large, regional power grids to be migraine inducing at best.

    With HVDC, you just connected them. Job done! In fact, HVDC lines are used heavily over much shorter distances where normally it wouldn't make sense to use a HVDC line, because a HVDC line allows you to connect two HVAC grids together without any concern over their frequency or phase. HVDC rectifies AC to DC, and later inverts it back to AC, and so only one end needs to be synchronized, and with the very grid it is connected to. Just look at this map of HVDC lines in Europe:

    enter image description here

    (Red = currently active, Green = under construction, Blue = proposed).

The downside is that the conversion and active, solid-state step up/step down switching facilities are expensive and, at least for now, not quite as reliable as the passive transformers that perform the same task for HVAC. This is the primary reason that HVDC, while more efficient even at shorter distances, does not have enough of an edge to offset the higher costs of building and maintaining HVDC step-up and step-down facilities. The technology is still maturing, which also means reliability will be not as good, and all of that makes HVAC the clear winner for still for power distribution, all the way to your home or business, at least at this point in time. The nice thing to remember is that power companies are not too keen on wasting power, so as new technologies emerge, they'll be adopted when and where it will make sense to do so.

  • \$\begingroup\$ good job ....but my half ass comment still stands from above....if you can't do anything with it once it gets there then AC is the clear choice. Now for the real question. If you take x amount of mechanical motion and convert to electricity using AC in one power station and the same x amount of mechanical motion and convert it to DC power and transmit each as is undisturbed or rectified then after they get there step them down to y number of units of equal work ..who wins ? \$\endgroup\$
    – Sedumjoy
    May 28, 2017 at 20:10

HVDC design engineer by trade here.

Many neat long answers with lots of information about cable costs, corona and inductance but the OP asked about what is the most (energy) efficient.

The inductive "loss" in an AC transmission is not a loss over the inductor itself but rather consequences of it which needs to be taken account for. If you have high enough system voltage to design with, this isn't a big issue over just 500 miles. Problem is that you don't. Cost of everything rises considerably when you go up in voltage on your investment. Running costs isn't too much affected. Some countries even have fixed limits or steps in system voltage you need to adhere to and you may run out of headroom here. With costs aside and from an energy efficiency standpoint, your 500 miles AC transmission will work admirably well.

If you opt for HVDC, your cable/overhead line losses will be a bit lower for the same amount of money spent on the equivalent AC dito, but the semiconductor losses which may only account for 0.2-0.8 % per station may not sound so much but compared to the AC solution, this is massive. On top of that comes multi kW pumping losses and fan-fed cooling system for the water cooling of said semiconductots. Also, your HVDC stations will inevitably need transformers which can take the full DC voltage though their isolation which makes them much more complicated, expensive and lossy than their AC counterparts. Again, still in the 99+ % efficiency range but you AC cable might need no transformers on any end if you have the same voltage in both ends for feed-in. Even if you didn't, the equivalent AC transformers with the same power rating would be more efficient.

So when do you need HVDC? Even longer distances. Almost any underwater/ground cable in the double digit mile/km range. Asynchronous grids.


Theoretically DC is more efficient than AC, because AC current figths not only the DC resistance of the wire but also the AC reactance - the impact of the wire-to-wire capacitance and wire inductance.

However it is very hard (and in the past was impossible) to boost the DC voltage to hundreds of kilovolts, what is fairly easy with AC using a transformer. Carrying power at a voltage of 2kV DC would require 100 times the current as if it was 200kV AC for the same power. And this is the key, because the losses are proportional to the square of the current.

For a given voltage DC is more efficient, but AC can be easily transformed to much higher voltage where is its best advantage.

  • \$\begingroup\$ Edison gets the last laugh...he should have worked on figuring out how to step it up and down.....didn't they have transformers back in the day for DC..???? I though they went that option? \$\endgroup\$
    – Sedumjoy
    May 27, 2017 at 20:17
  • 6
    \$\begingroup\$ @Sedumjoy Nowadays we still don't have transformers for DC. We have converters. There is a big difference ;) \$\endgroup\$ May 27, 2017 at 20:18
  • 2
    \$\begingroup\$ The other problem with AC is the skin effect. Wire diameter is somewhat limited by skin effect. At 60 Hz, skin depth is 8.4mm, and at 50 Hz, skin depth is 9.2mm. \$\endgroup\$
    – user57037
    May 27, 2017 at 20:29
  • \$\begingroup\$ @mkeith It's higher for aluminium and if you are using centimeter thick conductors, you need to go duplex anyway for mounting and mechanical reasons in most cases. You rarley run in to skin depth problems for overhead lines. \$\endgroup\$
    – winny
    May 31, 2017 at 7:01
  • \$\begingroup\$ @winny, maybe so. But I think it is worth mentioning in the context of this question. \$\endgroup\$
    – user57037
    Jun 1, 2017 at 2:27

DC has no skin effect thus less loss, but does have more serious PD effects that can lead to avalanche. So why do you think they don't use DC for distribution? and only for long haul transmission.

Reactance is not lossy but can be a PITA with reclosures. Why?

Which is more expensive? When is it cheaper?

  • \$\begingroup\$ I see your point....I am wondering about the article in Scientific American. Isn't 500 mile long haul? The article says the DC is vastly more efficient. The article in my view contradicts what I have heard...but maybe my old belief is just plan wrong.? \$\endgroup\$
    – Sedumjoy
    May 27, 2017 at 21:51
  • 1
    \$\begingroup\$ Anything more than 200kM to a grid node with multiple sources is possibly worth considering for DC. One simply get lower transmission losses but maybe a 2% more conversion losses. But no skin effect losses are significant. \$\endgroup\$ May 28, 2017 at 3:13

The article must not have mentioned the issue with AC peak voltage versus Vrms, line corona and arcing, and the higher cost of lines above one megavolt.

If a line is designed to tolerate 1MV at 100amps, then as a DC line it transmits 100MW. But for AC, the AC peak voltage must remain at one megavolt (since an AC operating voltage of 1MVrms would have 1.41MV peak, and the 1MV transmission line would break down, or at least display massive corona loss.) So, our AC line can only tolerate 100MV peaks, for 70.7MV rms at 100amps, only giving 70.7 megawatts versus DC's 100MW.

I've been hearing that because of expanding population and power demands, many main trunks have hit their limits, and must now operate at capacity. If it's cheaper to convert an existing trunk to DC, as opposed to stringing entire new rows of towers alongside the existing ones, then HVDC is going to be extremely popular with power companies.

So, it's not a matter of efficiency. It's a matter of corona losses imposing a voltage limit on existing lines, and the fact that HVDC lacks all those 120Hz voltage-peaks which exceed the corona limits by ~40%.


The keyword for HVDC step-up and down is: valve hall. "Valve," as in the British name for vacuum tube. I'm told that original DC intertie used yards-long mercury thyratrons. Today they use hockey-puck thyristors, in cubical arrays suspended from the ceiling.

  • \$\begingroup\$ Isn't it also a matter of reducing capacitive and inductive losses? \$\endgroup\$ May 28, 2017 at 10:06
  • \$\begingroup\$ @immibis maybe, if those losses were expensive or significant, like several tens of percent of system wattage. Aren't they well below 1%? I've only seen articles about losses to corona-breakdown and cable resistance, and about that enormous increase in throughput made possible by converting existing HVAC lines to HVDC. Still it's all economics: if the money lost in wasting power is much smaller than profits gained in overloading your network (making it inefficient,) then overloading your network is the right business decision. Since AC overloads at a much lower voltage, DC is superior. \$\endgroup\$
    – wbeaty
    May 31, 2017 at 6:00

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.