This group brings together the best thinkers on energy and climate. Join us for smart, insightful posts and conversations about where the energy industry is and where it is going.


AC/DC: In the New Current Wars, Will Edison Win Out After All?


In the late 1800s, as the United States embarked on a public effort to electrify the country, the “War of Currents” pitted Alternating Current (AC) power against Direct Current (DC) power. On one side was Thomas Edison, an advocate of DC power.  On the other side was George Westinghouse, who had acquired the patents for AC power held by Nikola Tesla. In the end, the capability of AC power to run at high voltage from large central generating stations over long distances with relatively inexpensive transformers to “step down” and “step up” the voltage made AC the victor, setting the industry standard. As the electric power sector modernizes, however, two substantial changes could signal a return to the war of currents.

First, declining costs and increasing options in distributed generation (DG) may offer an opportunity for microgrids and low-voltage transmission close to loads. These are the optimum conditions for the utilization of DC power, which can be more efficient than AC.

Second, improvements in battery technologies are driving down costs for localized energy storage. Batteries, which deliver DC power, offer many advantages to the electric system for supply side management, but also for demand side management. Eliminating spikes in demand could dramatically reduce expensive peak power generation needs. 

Combining storage with distributed generation, DC-powered microgrids become a very attractive option for their system efficiency, reliability, resiliency, and ancillary benefits.

The truth is, a surprising amount of today’s electronic devices are already DC. When I plug in my laptop, what goes into the computer itself is DC power – the power cord contains an inverter to change the current from AC at the plug to DC at the device. Every smartphone or other electronic device that charges from a USB cable uses DC power. That little USB plug that you put in the wall – that’s an inverter.

Variable speed motors, which are much more efficient than fixed speed motors, contain DC inverters, because DC offers the flexibility to meet varying speed requirements.

LED lights are also based on DC technology. The bulb contains a micro-inverter to change from AC to DC power, but at a cost. The conversion leads not only to reduction in efficiency, but also reduction in operable life. LEDs already last longer than incandescents and CFLs, but when they fail, it is the usually inverter that goes, not the bulb itself.

The biggest opportunity comes from the fact that the leading form of distributed generation, solar photovoltaic, creates electricity in DC. Now, solar owners need an inverter to get the power into AC for the grid and our AC-wired homes – again, at a cost. The typical derating of solar from nameplate capacity, from a variety of causes, is 23%, with roughly 8% of the reduction due to DC to AC conversion. Furthermore, solar systems that feed directly into a DC storage system don’t require net metering, which has come under attack as solar installations have grown.

ac-dc-logo-stencilSo, if much of our consumption requires DC power and distributed generation technologies generate DC, why should we lose so much energy by converting the current to AC and then back to DC? What if we turned the system around, inverting only where AC is required instead of DC? From an energy standpoint, that may be the future.

At the Devil’s Thumb Ranch, an eco-resort in the mountains of Colorado, Schneider Electric teamed up with advanced energy engineering firm PosEn to transform almost 200,000 square feet of buildings from standard AC to a DC microgrid system with lithium-ion battery storage and intelligent controls, along with distributed generation.

At a new barn and veterinary facility built to use only DC power, the system powers a well pump, motors, lighting, and trough heaters in a harsh environment (it gets down to -40 degrees F).  The power consumption of the facility was originally designed for about 10kW of AC – but that dropped to just 3.2kW utilizing locally generated DC power, sophisticated energy management, and energy storage.

There are other advantages to low voltage DC as well. Take the laptop power cord.  If you were to cut the cord and grab the wires on the AC side of the adapter, you would get a nasty shock. You would also create a fire danger. But on the DC side, cutting the cord would not cause a fire, and the shock would be much less. (Still, don’t try this at home). Wider utilization of low voltage DC would also make electrical work safer and easier. The National Electric Code (NEC), that is the standard across the country requires a licensed electrician to do work involving AC power – but low voltage DC can be done safely by a knowledgeable layperson, reducing construction and remodeling costs. 

Or take local distribution lines. Those wires that run from pole to pole are expensive to maintain and are exposed to the risk of falling tree limbs in storms, but the cost of putting them under ground is prohibitive. Low voltage DC lines can be undergrounded for much lower cost – giving the community a highly resilient distribution system without unsightly wires.

With all of these benefits of DC in a changing technology landscape where on-site DC generation and DC-powered devices are both growing in popularity, we could see a return to the current wars – and this time, Edison just may win.

AC/DC image source: Flickr. Used under a Creative Commons license.

Tom Plant's picture

Thank Tom for the Post!

Energy Central contributors share their experience and insights for the benefit of other Members (like you). Please show them your appreciation by leaving a comment, 'liking' this post, or following this Member.


Bas Gresnigt's picture
Bas Gresnigt on Feb 3, 2015 9:08 pm GMT

For the high capacity long distance backbone lines, high voltage DC is already more efficient than AC.
With the upcoming high efficient DC-DC converters, DC becomes gradually efficient in lower capacity lower voltage parts of the grid!

Thanks to the semiconductor developments, those DC–DC converters can replace the role of AC transformers. They may in the end even be less bulky than AC transformers (converting high-voltage DC to low-voltage DC and reverse).

Bob Meinetz's picture
Bob Meinetz on Feb 3, 2015 8:08 pm GMT

Tom, you ask

So, if much of our consumption requires DC power and distributed generation technologies generate DC, why should we lose so much energy by converting the current to AC and then back to DC? What if we turned the system around, inverting only where AC is required instead of DC? From an energy standpoint, that may be the future.

AC permits transforming electricity to higher voltages with less waste (99%+ efficiency) compared to DC (75-95% efficiency), and higher voltages transfer energy more efficiently through wire, whether the current is AC or DC. Unless the components of your computer, flatscreen TV, or LED bulb require the exact voltages your DC source is generating they will have to be converted to the correct ones, possibly several times.

Some applications with large energy requirements but fewer voltage conversion steps (data centers) are ideally suited to DC microgrids. Residential/industrial customers with a variety of equipment will continue to rely on AC because it is efficiently converted to various voltages at the point of use.

Whether inverting DC power in a microgrid is worth the inefficiency of conversion depends on the system, but it will be continue to be necessary for connections to the grid.

Clayton Handleman's picture
Clayton Handleman on Feb 3, 2015 7:55 pm GMT

We live in a time where we can choose.  Conversion from AC to DC and back is very efficient and inexpensive so we can move towards technology where the optimal source is used.  For example, in long distance transmission, it is more efficient to use High Voltage DC and less costly as the power lines exceed a few hundred miles.  The ohmic losses are less due to higher voltages and elimination of the corona and skin effects.  For underwater transmission, losses through capacitive coupling are eliminated.

However AC is probably easier and less expensive for shorter distance transmission as well as distribution and the risk of fire is far lower with AC.

In household devices, the DC voltages vary so much that a custom converter is often going to be required and there is little difference in efficiency between AC-DC and DC-DC conversion. 

Up conversion and Down conversion are pretty much a wash efficiency wise as far as AC transformers vs Solid State converters go.  This is in evidence when looking at PV inverters that now are routinely above 95% efficiency and as high as 98.5% efficient. 

This flexibility will only increase.  Bottom line, it is not an either or or a who wins kind of situation.  It is more a situation with 50 shades of gray and there will arise a growing diversity of approaches.

Roger Arnold's picture
Roger Arnold on Feb 4, 2015 9:22 am GMT

If you look at just the transformer unit itself, operating with a near-unity power factor, then utility AC transformers probably do still hold an edge over high voltage DC – DC converters. But it’s the difference between 98.5 – 99% vs 99.5%. Not a big deal, and DC – DC is definitely smaller and (these days) cheaper.

I don’t know if it’s true, but I recall hearing that you can’t even get a utility size high voltage AC transformer made in this country anymore. They have to be special ordered from one of the few overseas companies that still builds them, and wait times for delivery are over a year. That’s hearsay, and if someone else could confirm or refute it, I’d be interested in knowing.

In any case, the real difference in efficiency shows up at the system level, and strongly favors DC. It’s not just the higher average voltage leading to lower line losses, but the elimination of reactive power. AC transformers in practice don’t usually operate at 99.5% efficiency, because current and voltage are rarely in close phase. You get a lot of current flowing through the lines at portions of the AC cycle when the voltage is close to zero. That wastes energy. Utilities use capacitor banks, inductors, and synchronous capacitors distributed around the system to manage the problem and keep current and voltage from getting too far out of phase, but it makes for a system that’s a far cry from KISS.

A DC grid would be inherently more resilient. There’s no requirement for phase locking the multitude of generators operating on the grid, and no possibility of reactive power surges and loss of frequency sync taking one generator after another off line in a cascading blackout.

Unfortunately, we won’t be seeing that on a wide scale any time soon. Too much inertia with the existing AC infrastructure.

Roger Arnold's picture
Roger Arnold on Feb 4, 2015 9:33 am GMT

Note, however, that in equipment using high efficiency AC motors, the motors achieve high efficiency via integrated power factor correction boxes on their electrical feed. What’s a power factor correction box? Well, it can be viewed as a specialized form of AC to DC to AC converter!

Bas Gresnigt's picture
Bas Gresnigt on Feb 4, 2015 10:03 am GMT

We are still far off the end of semiconductor developments*).

DC-DC converters will become more efficient and may surpass that of AC transformers. Note further that AC transformers are bulky, heavy and require lot of expensive copper.
So DC-DC converters may become cheaper, more easy to handle and will be more flexible (thanks to their electronics).

*) We are near the end of developments with AC transformers (unless unexpected breakthroughs). Just as with airliners, etc.

Bas Gresnigt's picture
Bas Gresnigt on Feb 4, 2015 9:54 am GMT

Why is the risk of fire lower with AC?

Rick Engebretson's picture
Rick Engebretson on Feb 4, 2015 11:27 am GMT

I entirely agree that distributed DC microgrids are the game changer in electric power. All the elements exist in automotive systems. It has nothing to do with utility scale grids.

It is a non-disruptive evolutionary step much like the local smart personal computer began doing all the graphical internet browser work, while the slow telephone grid internet fed kilobytes of data reliably. Communications networks are in fact “routed” to ever smaller network grids.

If you want windmills and solar panels with batteries running your own microgrid; wonderful.

James Van Damme's picture
James Van Damme on Feb 4, 2015 5:00 pm GMT

I guess he’s referring to the nasty arc you get when you disconnect a big inductive load. With AC, it quenches faster because the voltage goes to zero within 16 msec. Like all generalizations, of course, “it depends”.

James Van Damme's picture
James Van Damme on Feb 4, 2015 5:07 pm GMT

Mr. Plant’s explanations about the laptop supply and underground distribution point out that government policy needs to be written by engineers, not politicians.

Bob Meinetz's picture
Bob Meinetz on Feb 4, 2015 5:52 pm GMT

James, an anecdote:

Back in the days when you could get away with this crazy kind of stuff (~1970) I had an eighth-grade science teacher who was supportive of kids following wherever their curiosity might lead them. Having recently seen a searchlight working up close, I decided my class project would be building a carbon-arc lamp.

I connected two leads from a wall socket to two copper tubes mounted in a wooden stand, into which thick artists’ charcoal pencils, flattened on the end, had been pushed. By very carefully tickling one pencil again the other then pulling it away, I was able to generate a spectacular, continuous, AC-powered carbon arc. The smoking and sputtering arc was too bright to look at directly, but using a magnifying glass I projected it onto a wall and could see the swirling 3000°C plasma between the pencils.

Your analysis sounds correct so I don’t understand why this even worked, but I got an “A” on the project. I never told my teacher about the time I got distracted and briefly used both hands to try to bring the pencils together, when I learned an important lesson about “short circuits”.

Disclaimer: do not try this at home.

Bob Meinetz's picture
Bob Meinetz on Feb 4, 2015 6:26 pm GMT

Roger, with the inefficiencies and other problems inherent in DC generators, how are we going to efficiently generate enough DC to power a grid?

James Van Damme's picture
James Van Damme on Feb 4, 2015 6:30 pm GMT

Actually, I have a carbon arc torch that does the same thing. It hooks up to my cheap “buzz box” welder, which is a step-down transformer with a high inductance which keeps the arc going. You need to start the arc, then the hot gas reignites the arc.

James Van Damme's picture
James Van Damme on Feb 4, 2015 6:42 pm GMT

The same way you do it in your car: generate AC, and rectify it. Using thyristors, you can control the output.

Bob Meinetz's picture
Bob Meinetz on Feb 4, 2015 8:00 pm GMT

James, isn’t that considerably less efficient than AC generation?

James Van Damme's picture
James Van Damme on Feb 4, 2015 8:16 pm GMT

DC generators used commutators to get the phasing right so you get DC out. They are a maintenance problem. With AC you just have 2 constant field winding leads, and the resultant AC is sorted out with diodes.  Chrysler developed alternators in the early 60’s, and they’re universal now.

Utility AC generators have to be matched in frequency and voltage, which is tricky when you’re coupled to an engine of some sort. If you rectify to DC, you can feed an inverter at whatever voltage and frequency the grid wants. You can also store the DC in batteries, which isn’t common yet, but it will come to that someday with solar PV and more windmills.

Switching semiconductors, like your light dimmers, have low losses relative to the megawatts they wheel around.

Bob Meinetz's picture
Bob Meinetz on Feb 4, 2015 9:19 pm GMT

James, in the U.S. AC utility generators are geared physically to generate 60Hz. The generator rotors are not spinning at 3600rpm, but the coils are configured to match some divisor of that (4 coils at 900rpm, for example). It’s not particularly tricky to get them up to speed and maintain them in phase – in modern combined cycle gas turbines it happens in under a minute.

Bob Meinetz's picture
Bob Meinetz on Feb 4, 2015 9:13 pm GMT

Roger, I’m not understanding your comment

You get a lot of current flowing through the lines at portions of the AC cycle when the voltage is close to zero…

When the voltage is close to zero, there’s proportionally close to zero current flowing through the lines (E = IR).


Roger Arnold's picture
Roger Arnold on Feb 4, 2015 10:01 pm GMT

Go back to Maxwell’s equations. Current flows in response to an electric field — that is a voltage gradient. There’s no direct relation to the “absolute” voltage (with respect to ground), except to the extent that if one end of a wire is grounded, an applied voltage at the other end will certainly set up a voltage gradient withing the wire and cause current to flow.

When you use “E = IR”, you’re implicitly operating in a DC universe. Analagous to special relativity, which excludes acceleration and gravitation. In an AC universe, you have inductance and capacitance to worry about. Bottom line: yes, one certainly can have a large current flow at zero voltage (at the point at which one happens to be measuring the voltage). That’s what reactive power is about. In the extreme case — a  purely capacitive or a purely inductive load, current and voltage are 90 degrees out of phase. No net power is being delivered to the load at all, but lots of current is flowing back and forth, and energy is being dissipated like crazy.

Bob Meinetz's picture
Bob Meinetz on Feb 4, 2015 10:46 pm GMT

Thanks for explanation.

Clayton Handleman's picture
Clayton Handleman on Feb 5, 2015 2:36 am GMT

Boost converters do not require any AC to AC stage.  No passive transformer is requried.  The passive is an inductor.

The power semis vary depending upon the job they are being asked to do.  From MOSFETs to IGBTs to others.  The semis are very low loss when on or off, much of the loss comes from the switching phase.  The faster they go through the transition the less loss.  Much of the design effor goes into managing the very high frequency harmonic content associated with the fast edges of the semis turning on and off.  

As far as cost and reliability, complexity is a non-issue.   Cell phones are complex but inexpensive.  PCs are complex but inexpensive.  Once designed, things can be manufactured inexpensively and built reliably. 

Great example is inverters for grid tie.  Very complex, lots of engineering, relatively inexpensive, very robust.

Nathan Wilson's picture
Nathan Wilson on Feb 5, 2015 5:30 am GMT

Tom, I think you are mis-attributing the safety of DC appliances like user-installed yard lights and laptop power cords to the use of DC, when in realty, the safety is entirely due to the low voltage (12 V, normally AC for yard lights) and low power limit that is built-in.

Your car battery is also 12 V, just try hooking a wire across the terminals (don’t do this at home): you can expect to be sprayed in the face with hot-flamming molten copper in addition to having the wire burned-out of your hand.  This differs from your experience cutting the lap-top cord because the laptop’s power supply is internally limited to 100 Watts or so, but the car battery can put out a few thousand Watts for a short time.

Similarly, the Schneider Electric system (such as the Devil’s Thumb Ranch application you mentioned) is intended for micro-grids in which each user needs only a small amount of power.  So it distributes power as 230 VDC, then reduces it to 14 VDC for each user.  The Schneider product description boasts that the system owner’s equipment is very reliable, as the 230-14V power converter and the batteries are on the customer side, which means they don’t count against the owner’s reliability rating (but of course they impacts users).

But while low power (<100 Watts) is a step up for third world villages with no power, normal 120/240 VAC grid power is much more versatile still, as it allows appliances such as air-conditioning, clothes dryers, and all-electric kitchens, (which take many KWatts; US houses typically have 100A @ 120/240V service, which gives a power limit of 24 KWatts).  These high power levels could also be provided with 120 VDC, but there is little incentive to do this, since most electricity users live in places where being on-grid is cheaper than being off-grid; even off-grid, as Clayton points out, the cost of the inverter is low and falling.  Changing from one standard to another is easiest when the new standard gives exciting new features (like Rick’s example of PCs and phone-line internet), but changing a standard to get 10% more efficiency is a tough sell.

I think one of the most telling trends in micro-grids is the data center example.  In the 1970s, telephone company switch offices were among the first battery-backed micro-grids, and they ran on 48 VDC power.  Now modern data centers need the same battery-backed reliability, but the power needs are much greater, with each rack of equipment needing many kWatts.  As a result of the higher power, 120 V and 240 V is a more appropriate solution.  But they don’t use 120 or 240 VDC, since the cost of the inverter is low compared to the batteries and control circuitry, plus compatibility with the standard grid AC voltage is a benefit.

Clayton Handleman's picture
Clayton Handleman on Feb 7, 2015 3:43 am GMT

The AC zero crossing often extinguishes the arc.  A friend of mine was reparing a solar array and forgot to cover it or pull the fuse.  The wire was dual conductor and when he pulled the connecter apart a spark jumped, ionized the gas in the immediate vicinity and an arc occurred.  It ran up the line until it got into the diode box where the conductors separated enough to extinguish the arc.

Clayton Handleman's picture
Clayton Handleman on Feb 7, 2015 3:44 am GMT


Nathan Wilson's picture
Nathan Wilson on Feb 7, 2015 3:38 pm GMT

sun panels and a small wind turbineShe would benefit greatly from not having to use an invertor to make all her power AC

No, she’d only save around 10% on the cost of her system (note that she still needs a power conditioner box to manage her battery).  But the down side is that she’d likely pay 10-100% more for all of her appliance to get specialty AC/DC capability, plus there would be much lower selection available with that feature.

It is hard to bring out a new inter-operability standard for commodities just to get a small efficiency improvement.  “Upgrades” require exciting new features, which DC does not offer.

Bruce McFarling's picture
Bruce McFarling on Feb 10, 2015 3:09 am GMT

I can’t see how DC to DC conversion will ever be as efficient as AC to AC conversion.”

However, the suggestion there is that if long distance (or buried) transmission is via HVDC lines, then the design problem is not DC to DC conversion versus AC to AC conversion, but rather DC to DC vs DC to AC.

I’d still suspect that due to the multi-drop problem with DC transmission that an AC remains the solution for the primary distribution grid. That means that even if there is an alternate HVDC supply link, the micro-grid could be DC, with an AC to DC conversion to the primary distribution grid and a DC to DC conversion to the alternate source of supply, or AC, connecting to the AC distribution grid via transformer with a DC to AC conversion to an alternate source of supply.

Get Published - Build a Following

The Energy Central Power Industry Network is based on one core idea - power industry professionals helping each other and advancing the industry by sharing and learning from each other.

If you have an experience or insight to share or have learned something from a conference or seminar, your peers and colleagues on Energy Central want to hear about it. It's also easy to share a link to an article you've liked or an industry resource that you think would be helpful.

                 Learn more about posting on Energy Central »