We know that we have now 50/60Hz in our walls due to mainly historical reasons - back 100 years ago there were no ways to up/down scale DC voltage.

These days we just have problems due to that - every single device sold have to have ~1uF cap per 1W of power before it's PSU to have enough power when we go through 0. (this problem does not exist in 3-phase power, but it available mainly in industrial applications only AFAIK) + caps have to have higher rated voltage to survive sine peaks + all this PFC mess.

Is that correct to say that if we were to design modern power grid, we would skip AC, and just have DC everywhere? As far as I see, it would significantly increase reliability & reduce cost of many devices out there.

  • 10
    \$\begingroup\$ @Leon Heller I am really starting to get annoyed with how short you are with things on this site. It really is not needed. If you don't like something you need to explain yourself. \$\endgroup\$
    – Kellenjb
    Commented Feb 27, 2011 at 16:36
  • 4
    \$\begingroup\$ Another idea is to have a centralized, well-designed SMPS for each house, and supply a few standardized DC voltages to special outlets, so you aren't wasting copper and energy on tons of inefficient wall warts and brick adapters. \$\endgroup\$
    – endolith
    Commented Feb 27, 2011 at 23:43
  • 2
    \$\begingroup\$ @endolith that is the idea that I have always loved. The logistics of changing a grid to DC are difficult regardless of technical difficulties. Keeping our existing infrastructure and just distributing the SMPS will be the cheapest route. No reason houses can't implement this now. \$\endgroup\$
    – Kellenjb
    Commented Feb 28, 2011 at 2:01

7 Answers 7


Guy Allee at Intel Research wrote about this topic last year -- DC - An idea whose time has come and gone? -- in support of a 380VDC grid, with the following bullet points:

  • 7% Energy Savings vs. High-Efficiency 415VAC; 28% vs Current Typical 208VAC
  • 15% Less Capital Cost
  • 15% fewer PSU components
  • 33% Datacenter Space Savings
  • 200% Reliability Improvement, which goes to 1000% if you directly connect the battery bus
  • Elimination of harmonics and inherently immune to other AC power quality issues
  • Natural affinity to alternate energy generation (Photovoltaic, and wind are ~400Vdc internally, and you actually lose energy & efficiency when you are forced to convert to AC)

He added in the comments:

We very deliberately picked 380Vdc because you want to get to as high a voltage as you can afford for efficiency. At the same time this standard is targeting Low Voltage applications only (<600V). We would have gone higher, but there are structural cost barriers at 400Vdc and 420Vdc. At 380Vdc we stay with the same volume parts ratings that AC is using and get the volume cost benefits of piggybacking on the bulk of current AC power supply component volumes. I’m sure you can also appreciate the significant cost adders that +/-340Vdc has on the personal safety equipment, which is why the standard allows for a cost-effective +/-190Vdc distribution. Thus we have the highest efficiency yet cost effective standard. And with the affinity among other industries, PV, wind, electric vehicles, and lighting, the volume economics seem compelling.

He also mentions the idea of a mixed distribution of both AC and DC within a building (e.g. data centers). For more on that initiative, see the EMerge Alliance website: http://www.emergealliance.org.

  • \$\begingroup\$ I wonder if it's feasible to have 380 at homes, must be too dangerous... \$\endgroup\$ Commented Feb 27, 2011 at 18:05
  • 2
    \$\begingroup\$ Well, we have 400 V AC 3-phase in just about all homes around here (Denmark), so it's certainly not any more dangerous than what we're already doing. \$\endgroup\$
    – dren.dk
    Commented Feb 27, 2011 at 18:27
  • \$\begingroup\$ Presumably he means 380VDC as the feeder to home. I wonder what his plans are for actual transmission/distribution. He certainly can do high voltage DC (probably 3k-200k V depending on the distance/load of the run) but has to come up with a cheap and efficient alternative for a pole transformer which would have to convert something on the order of 3k-30k VDC to 380VDC (assuming similar current levels as AC systems). \$\endgroup\$
    – Mark
    Commented Feb 28, 2011 at 6:47
  • \$\begingroup\$ Yes, it's a local grid of +/- 190V (with, e.g., 24V outlets in data centers). Here's a Wikipedia list of high voltage DC projects: en.wikipedia.org/wiki/List_of_HVDC_projects \$\endgroup\$
    – Eryk Sun
    Commented Feb 28, 2011 at 12:15

Safety. Having HVDC through the wall outlet is not smart. Unplug a high current device without first switching it off will pull a huge arc

  • 1
    \$\begingroup\$ Use a flyback diode. \$\endgroup\$
    – Eryk Sun
    Commented Feb 27, 2011 at 21:04
  • \$\begingroup\$ Just like in AC - if you pull the plug in the right time, you will be cutting 380v... \$\endgroup\$ Commented Feb 27, 2011 at 21:14
  • 5
    \$\begingroup\$ AC self extinguishes to zero 50/60 times a second though. You CANNOT switch HVDC in the same manner as AC. \$\endgroup\$ Commented Feb 27, 2011 at 21:39
  • 7
    \$\begingroup\$ @Eddie - It even self extinguishes 100/120 times per second! \$\endgroup\$
    – stevenvh
    Commented Jul 8, 2011 at 6:48

Short answer:


Long answer:

The advantage of AC for distributing power over a distance is due to the ease of changing voltages using a transformer. Converting DC power from one voltage to another requires a large spinning rotary converter or motor-generator set, which is difficult, expensive, inefficient, and required maintenance, whereas with AC the voltage can be changed with simple and efficient transformers that have no moving parts and require very little maintenance.

Suggested reading:

War of Currents

  • 1
    \$\begingroup\$ you also have to look at the power plants themselves. Most power plants create some sort of AC through mechanical means. Are there ways of converting That to DC efficiently, for DC power transmission at such high levels? \$\endgroup\$
    – jsolarski
    Commented Feb 27, 2011 at 16:57
  • 2
    \$\begingroup\$ @Dean, There are health risks with being near a substation? Do you mean getting electrocuted? \$\endgroup\$
    – Kortuk
    Commented Feb 27, 2011 at 19:01
  • 3
    \$\begingroup\$ @Andrejako, people believe alot of things, lets stick to science <3 \$\endgroup\$
    – Kortuk
    Commented Feb 27, 2011 at 20:25
  • 1
    \$\begingroup\$ @Kortuk the possibility of the magnetic fields causing damage to human health. I personally think its a load of rubbish. \$\endgroup\$
    – Dean
    Commented Feb 27, 2011 at 20:37
  • 3
    \$\begingroup\$ I think this answer is wrong. If you look up the HVDC page on Wikipedia, the disadvantages of DC transmission are that it has to be converted to AC. The reason AC was chosen over DC was that there was, at that time, no efficient means of stepping up and down voltages. With technology today, this is no longer a problem. New long distance power links these days are built using DC, as it is more efficient. \$\endgroup\$
    – Mas
    Commented Oct 9, 2012 at 19:00

You may be right. AC once held a huge advantage over DC in the past. But as the cost of DC-DC converters has dropped, the relative advantage of AC has dropped and in some cases crossed over. If we were designing a new power transmission system today, DC everywhere might reduce total system costs.

For an equivalent power and current levels and reliability, DC requires slightly stronger parts for circuit breakers and fuses and lightning arrestors; but AC requires slightly more expensive transmission lines and better coordination of power generators to avoid cascading failure.

Even though (for historical reasons) AC equipment has mass-production economy-of-scale advantages over DC equipment, the designers of many recent long-distance power transmission systems have apparently decided that using high-voltage DC (typically 200,000 VDC) has lower net system costs than using AC.

Even though (for historical reasons) many airplanes and the Space Shuttle use 400 Hz 120 VAC, early plans for the international space station called for it to use a 20,000 Hz 440 VAC distribution power (!), until program priorities changed and the engineers switched to 120 VDC power. (Mukund R. Patel p. 543)

People at Google (a,b) have suggested to desktop and server manufacturers that net cost could go down if we switch to "12V-only supplies" that convert the AC mains power to 12 VDC, and then the computer motherboard requires only 12 VDC, which it steps down to whatever collection of voltages it needs (like most laptops), rather than the current ATX power supply configuration that has a thick bundle of wires with a motley assortment of voltages.

Lee Felsenstein and Douglas Adams have gone even further and asked that someone develop a standard 12 VDC distribution system. (c,d)


There is another point, that i like to add, why we can't skip AC in my opinion. Long tracks, especially cables are better done in DC (because of the inductance/capactiance which are expensive to handle at longer distances.)

The big thing is, that HVDC-lines are point to point. An meshed-DC grid is a whole other story. If at any point of the grid an error occurs, e.g. a tree falls on the line, the whole meshed-net is down (the voltage drops to near zero, and the converters have to shut down).
In AC the impedanze mostly is influenced by inductance, so we have much bigger impedanze as in DC, where the impedanze euals the small resistance. If a tree falls into an AC-line the volatge at that point is zero. But the high error-current and the high impedance make a big voltage. So just this line is out, the others (if not very near) have (almost) their normal voltage. In DC the impedance is very small, so the volatge in the whole meshed grid dropes to near zero and not just one line but the whole net is down. Also you should know, that the balance of power productiona and consumtion in AC is done via frequenzy. In DC it is done via Voltage. This should make it obvious, that such a large problem with the voltage is not good at all.
If someone would want to tranfsport any significant power over this net with low volage or wants to make the volatge higher, very very large currents are needed, so large, that the lines would just melt. Therefore the converters shut down (blackout) and wait until the line is repaired and ready.


Short answer: Not so fast Longer: Solid state converters are pretty good. Long haul transmission has lots of advantages. Short haul probably still benefits from transformers.

  • \$\begingroup\$ I don't understand both 'Not so fast Longer' and what is say in the second part... Both solid state converters & transformers are good???? :-| \$\endgroup\$ Commented Feb 27, 2011 at 16:44

Extra Info: There are some DC power lines in the world. Take by example the HVDC line at Itaipu, it remains among the most important HVDC installations in the world. It's a 6300 MW line with 780 km of length.

  • 2
    \$\begingroup\$ It should probably be noted that this sort of HVDC installation is normally done for other reasons, namely charging the capacitance of extremely long power lines leads to large reactive currents, which cause resistive losses. It's not much of a problem for shorter power lines, but when you have extremely long-haul power lines which do not have taps for supplying power to locales along their length, it actually becomes cost effective. \$\endgroup\$ Commented Aug 29, 2013 at 9:10
  • \$\begingroup\$ The usual reason DC links are chosen is to subdivide regions of AC transmission to make them easier to manage (e.g the Pacific Intertie). Or between UK and France. They can interconnect regions with different frequencies or phases. \$\endgroup\$ Commented Mar 5, 2015 at 3:24

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.