Lately, I’ve been reading about the many advantages of HVDC transmission systems for long distance transmission, undersea links, and others. The historical reason of why AC was picked over DC was mostly due to the invention of the transformer, which allowed easy manipulation of AC voltage enabling high voltage transmission across long distances.

However, after the invention of mercury valves, thyristors, IGBTs and all these components that have made DC transmission feasible, I’ve been thinking that if we had a purely DC network we could get rid of all AC/DC rectifiers that we find in our electronic appliances. This could improve energy efficiency greatly and save tons of money in resources.

If we had a chance to start over, could a DC-based transmission system be the better1 choice, or would AC still come on top?

1: with better, I mean more energy efficient.

  • 2
    \$\begingroup\$ There's no clear answer to this, which makes it a "non-constructive" question. However, I'd point out that to achieve galvanic isolation (which would still be required, even in a DC-DC converter) requires transmitting an AC signal across a magnetic or capacitive interface. \$\endgroup\$
    – Dave Tweed
    Commented May 1, 2013 at 18:54
  • 1
    \$\begingroup\$ You'd still need voltage converters; replacing AC/DC rectifiers with DC/DC converters isn't a noticable improvement. Now, there might be benefits to house-level low-voltage DC (12 or 24V?) \$\endgroup\$
    – pjc50
    Commented May 1, 2013 at 18:58
  • 9
    \$\begingroup\$ Judging by upvotes, the community seems to like the question. But before answering, please remember, "We expect answers to be supported by facts, references, or specific expertise," and not "debate, arguments, or extended discussion." \$\endgroup\$
    – The Photon
    Commented May 1, 2013 at 19:01
  • 4
    \$\begingroup\$ The question clearly states a criteria - efficiency. One could entertain the alternative interpretation of cost efficiency rather than energy efficiency but given the name of the site the first would be the default assumption. The body of the question alludes to both, so good answers might weigh the differences. \$\endgroup\$ Commented May 1, 2013 at 20:42
  • 1
    \$\begingroup\$ @ChrisStratton, we should also try to answer the question in the text, "If we had a chance to start over, ... would AC still come on top?" --- which is probably more about cost efficiency than power efficiency. \$\endgroup\$
    – The Photon
    Commented May 1, 2013 at 20:49

4 Answers 4


A Bit of History

The suggestions behind this topic go against what a lot of electrical engineers have been taught since their first circuits course - that AC is better for power transmission. After all, in the "war of currents" in the late 1800's, it was Tesla who helped Westinghouse fight for AC, eventually defeating Edison's dreams of a DC empire.

The primary advantage of using AC over DC at this time was efficiency. It became increasingly easy to transform one AC voltage to another, especially when compared to the cost, difficulty, and inefficiency of converting one DC voltage to another at that time. According to Joule's First Law, the amount of power wasted as heat in the transmission lines is proportional to the current squared. Considering the transmission lines have a known, (basically) fixed resistance, then for a transmission of the same amount of power, much more is wasted in a low-voltage, high-current transmission than a high-voltage, low-current transmission. As stated, it was very impractical to convert DC voltages to a high enough level to over come the line loss when compared to the relative ease of transforming AC voltages.

As a side note, many places never fully switched from the original DC transmission systems to AC until the mid 20th century.

You can read all about the history here.

Enter Modern Electrical Design

This is not to say AC doesn't have it's own problems. The skin effect is one example of AC being less efficient than DC, but it still doesn't compensate for the above mentioned line losses. Another issue is that of corona discharge occurring at high transmission voltage levels. Over long distances, AC power also has stability issues. This IEEE article sites a few different distances, noting that the line reactance can be compensated for in distances up to 600 - 700 miles.

With modern implementations of mercury arc valves, thyristors, and IGBTs, and efficient means of DC voltage conversion, HVDC transmission is not only possible, but overcomes many of the problems faced with HVAC transmission. The overall transmission distance is much greater, and the mentioned AC effects are overcome. In addition, the cost associated with HVDC is less than HVAC, once a distance threshold has been crossed. This cost differential is discussed in detail in this paper which includes a breakdown of power substation cost as well. The cost is also discussed in the link provided by Jake in his answer.

The fact is that the current electrical infrastructure is based around AC power transmission. The vast majority of modern technology requires this type of power for proper operation, and had AC never been used I doubt we would have many of the technilogical "advancements" we known and love. Theoretically, using HVDC alone could prove to be more efficient, but to compensate for the difference in cost, a hybrid HVAC/HVDC system is the best solution, at least in this moment in human development.

  • \$\begingroup\$ So, if I understand correctly, the easyness of use of AC at the time helped electricity to flourish during the 20th century. However, HVDC seems to be better choice today once above the break-even distance from a cost perspective. Disgregarding cost contraints, does HVDC have any techincal shotcommings when compared to current AC systems though? \$\endgroup\$
    – Jota
    Commented May 1, 2013 at 21:23
  • 1
    \$\begingroup\$ @Oniros - Exactly. Asking what might we do if we could do it all over again is a pointless question because our present ideology and decision making processes are based on history. If that history were to change, so would our line of thinking. A better question would be how can we improve upon the current electrical grid design to improve both cost and energy efficiency, given our past and present technology. \$\endgroup\$ Commented May 1, 2013 at 21:27
  • 1
    \$\begingroup\$ @Oniros, I don't know enough to comment on the shortcomings of the current HVDC system. I would imagine there is something about it in one of the links presented in my answer. Specifically, the "disadvantages" section of this wiki page: en.wikipedia.org/wiki/High-voltage_direct_current \$\endgroup\$ Commented May 1, 2013 at 21:37
  • \$\begingroup\$ Very nice explanation \$\endgroup\$
    – MeTitus
    Commented Apr 16, 2020 at 10:43

I think a good candidate for consideration would be 3-phase AC to every outlet.

  • Efficient rugged transformers can be used for isolation and changing voltages
  • No mains frequency reservoir capacitors required for DC power supplies or VFD motor drives
  • Efficient three phase motors can be used directly where a single speed is possible
  • Less copper required for the same power transmission and losses.

Cost is extra conductors and connector pins.

  • 1
    \$\begingroup\$ I second this! I can think of so many useful applications just off the top of my head. \$\endgroup\$
    – anrieff
    Commented Apr 19, 2020 at 8:48
  • \$\begingroup\$ That would have been an interesting design choice - can't really compare costs since most everything would be designed very differently (altered supply & demand). Would require more wires going to each outlet, but they can be smaller gauge. DC rectification would be slightly more complex, but you'd get higher amplitude with less ripple as well. \$\endgroup\$ Commented Apr 20, 2020 at 17:20
  • \$\begingroup\$ I believe power transmission is typically 3 phase to the residential load stations, where it's divided to be single phase hot/neutral lines. So all of the negatives of the AC power transmission grid I mentioned are still in full effect. \$\endgroup\$ Commented Apr 20, 2020 at 17:22
  • \$\begingroup\$ @KurtE.Clothier In North America, it's 3-phase on major corridors but typically only single phase down residential urban streets and to rural properties. Pole pigs are single-phase step-down from eg. 14.4kV to hundreds of volts (120:120). High voltage DC works well for long distance and undersea transmission, of course, so one would mix the two as is done now. \$\endgroup\$ Commented Apr 20, 2020 at 18:47
  • \$\begingroup\$ @SpehroPefhany Thanks for the confirmation. That's what I was thinking. \$\endgroup\$ Commented Apr 22, 2020 at 13:13

With the cost of copper and other useful metals rising, and the environmental issues with insulating oils used in them, high power AC transformers are becoming very expensive, compared to their solid state equivalents. It's not only the cost of the windings, but the large metal cases, and transportation and installation costs associated with such large transformers.

A switch to high voltage DC transmission would probably be a wash in terms of efficiency, however it could be lower cost if we didn't already have our current infrastructure in place. Transformers would still be needed, but rather than low frequency power transformers we would couple semiconductor DC-DC converters with high frequency power transformers which can be much smaller (thus cheaper) for the same amount of power conversion capacity.

  • \$\begingroup\$ "environmental issues with insulating oils used in them..." - These days you can get vegetable-based insulating oil which is non-toxic and biodegradable, so this is less of an issue than it used to be. As a bonus, the vegetable oil is also less easily flammable than mineral oil, so transformers are less likely to explode into fireballs under fault conditions. \$\endgroup\$ Commented Sep 29, 2013 at 6:15

The question is a meaningful one, but implies a simplistic answer is possible. It's really not. The optimum solution can be deduced by the transmission and distribution strategies that are being implemented all over the world. Look at the Wikipedia article or many other articles that can be found. In our early years Tesla actually worked for Edison. Edison was such a SOB generally and was so nasty to Tesla specifically, Tesla quit and basically gave all of his patents to George Westinghouse. Tesla KNEW that long-distance transmission of DC was not feasible AT THAT TIME because of the inability of step-up, step-down of voltage. Westinghouse (using Tesla's technology) supported the building of the Adams AC hydroelectric power plant near Niagra falls. It supplied power to Niagra and also as far away as Buffalo. Opened in 1985. It was the first major power plant ~2MW to distribute over long distances in the world. G.E. management was taken by surprise and they basically put Edison in a box. This comes from internal G.E. documents. They now realized that AC transmission was so important that they hired the "Wizard of Schenectady", Charles Steinmetz, since they needed somebody who understood AC power. What would happen today in the US? Our power generation/transmission/consumption facilities would be constructed with engineering judgement, using mostly AC, with longer transmission lines being built for D.C. transmission. We'd probably have more DC lines than we have now, simply because we wouldn't already have a lot of AC lines that are currently occupying right-of-way. AC/DC convertors are still extensive, so there probably wouldn't be a lot of DC power consumed locally AT THIS TIME. AC induction motors are much more simple than any DC motor I know of - happy to be enlightened if this is not true. Also, there is a reason that 50 or 60 hertz was chosen for transmission frequencies. It requires more iron in the motor, but allows the motor to generate power at a lower rotational velocity. Airplane electrics are 400Hz AC and have smaller and LIGHTER generators and motors. They don't have the power, though. Big power requires heavy internal windings for both generators and motors. 50-60 Hertz was not a capricious decision. Big power generators turn at MUCH lower frequencies, since their internal windings weigh TONS. Large generators often have twenty poles or more (turn at 3HZ, sometimes less). So, distribution to industrial consumers would probably remain at 60HZ AC. What about at your home? Local distribution voltages have to be around 2000-4000 volts, as they are now. That's what goes INTO the transformer on your pole. DC would have to be the same, just for transmission economics. So that would mean that DC electricity would have to be inverted from (e.g.) +- 400 KV to (e.g. 115 KV) AC, transformed to (e.g. 34.5 KV) AC, rectified to (e.g +-1 KV) DC and delivered to the pole. This voltage step-down strategy, since it has to do with what voltages can can be economically transmitted over what distances. Even if everything was all DC, which is not economically feasible, the +- 1KV at the pole could be inverted to 125V AC (you'd also need to invert to 250v AC to get the higher voltage for your stove that you automatically get from the AC "Edison three wire distribution" technique that's used everywhere in the US. If we redesigned EVERYTHING to use 250V AC (like other countries) this last would be obviated. Understand that even small (household) DC motors are expensive and more complex than AC motors, though. We could use DC motors in home appliances, but the cost would rise and reliability would go down. So, although the answer is not simple, there is an answer. We'd just have to do a lot of engineering......

  • 2
    \$\begingroup\$ This really is a big wall of text - try to format it in a nice way, so that people coming here are motivated to read you answer. As it is, a lot of people will just skip reading you post, because this "wall" is completely overwhelming. \$\endgroup\$
    – jusaca
    Commented Apr 19, 2020 at 10:24

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.