Switching the mains to DC

[torginus]

Hi!
This is a bit of a pie in the sky thought, but I was just thinking about what would be the downsides/upsides of switching the mains voltage in our homes to DC (with the same voltage as our AC RMS)?
Almost every device in our homes either uses DC voltages internally, or doesn't care. I think the last holdouts that do require AC were devices that had motors in them, like air conditioners, washing machines and the like, but even these are being replaced with ones that have BLDC motors internally.

In fact, almost all devices sold currently use switching power supplies, which have rectifiers on their inputs to create high voltage DC to step down, and as far as I can tell, those would continue to just work from either DC or AC, reducing the logistical pain for consumers and manufacturers. On the infrastructure level, the switch could be carried out by installing an AC-DC converter at any point, from the local grid, to apartment, so this doesn't need to be an all or nothing effort.

As for why, it's very simple - while this change would've made little practical sense years ago, now with the emergence of huge DC sinks (EVs) and sources (solar panels), I think having people purchase expensive and failure prone solar inverters is just bonkers. Solar panels are dirt cheap and if they could directly feed into the home circuit with just a voltage regulator, it'd change the economics massively, so I think there's cause to try this.

The only downsides I can see is that DC safety equipment (breakers, RCDs) is more expensive than the AC equivalents.

What do you think? Am I a bumbling idiot that has overlooked some fundamental flaws or is the idea sound and feasible? If it is, are there any pilot projects or proposals to do this?

[bdunham7]

What do you think? Am I a bumbling idiot that has overlooked some fundamental flaws or is the idea sound and feasible? If it is, are there any pilot projects or proposals to do this?

Well, there's the installed base of billions of induction motors, mains transformers and so forth that would become unusable.  It is trivially easy to generate DC voltage from AC nowadays with PFC supplies so you really aren't gaining much--those new devices that work better with DC can obtain it easily enough.  I don't think solar inverters are particularly unreliable and if they are, making them more robust surely can't be an insurmountable task.  More than 90% of my stuff absolutely requires AC unless I do substantial modifications.  I do have an old AC/DC radio, but I converted that to AC only input with DC internals using an isolation transformer and a HV linear PSU setup. 

[Circlotron]

Switches arcing.
Electrolysis.
More fuel for people that complain about 5G, smart meters, flat earth, and moon landing conspiracies.

[Psi]

It might make sense in the far future if all motors move to brushless DC and if semiconductors for working with DC mains voltages at low Rds become cheaper than mechanical switches.
As others have said there are mechanically switching problems with DC so that would all have to go.

Electrolysis is an interesting one that I'd not considered, yeah would have to deal with that somehow.

[coppercone2]

don't do this. The code is meant for 60Hz current. The behavior of plasma changes drastically at DC and basically requires special components to achieve the same performance levels as AC components.

60Hz arcing likes to extinguish itself. DC does not. Because of this all switches take a beating and over heating.

I am not sure how they feel about the insulation performance either (how much of a problem can it take at DC vs AC before it goes haywire.

Same with the outlets (devices that spark when you plug them in).

IMO you would need alot of testing to certify all that stuff. Look at how solar DC stuff is built and you will see how high current DC  is different then AC.


The corrosion issue is true too (the phone company has experience with this).




Its NOT a little difference. Also there is the issue of inrush with bus capacitors. In his little demo the arc breaking distance looks to be... 15 times longer

[David Hess]

Various safety devices are rated for AC; switches and breakers and other things would need to be upgraded for DC operation.

Based on my experience with ECMs (electronically commutated motors) which have replaced induction motors, they will be more expensive, and less reliable.

[Siwastaja]

Biggest issue is switches and fuses, all incompatible, replacements 100x more expensive, and missing to replace any single one results in house fire.

[nctnico]

Another problem with DC is that most loads are not constant but pulsating. So you'd still need the DC variant of power factor correction otherwise the 'DC' grid will probably start oscillating wildly due to impedance mismatches. In the end DC doesn't exist in the real world. 48V telecom equipment is a good example. There are strict rules for the frequency content of the current being drawn from the source.

[Seekonk]

There was a coalition of companies that were trying to create a new standard for low voltage DC in the homes like a 30V range USB to power DC products.  There would be one main converter for AC to DC instead of all these wall warts. Probably adaptable to solar too.  Haven't heard anything about that in years.

I have an off grid home and many of my inverters are DC out to run things. Much quieter and efficient.

[TimFox]

I believe that the highest loads on the power to my house are AC induction motors:
  Refrigerator
  Freezer
  Central air conditioner compressor.
  Forced-air heating fans on gas-fired furnace.
  Vacuum cleaner.
  Clothes washing machine.
  Motor for gas-fired clothes dryer.

What advantage to me is the expensive replacement of all of these to accommodate DC mains?
Transformers may be heavy, but they can be extremely reliable compared to anything using solid-state power electronics.

[Marco]

Switches arcing.

You can use a two pole switch and some passive components. DC switch gets protected during turn off by a parallel capacitor, which then resonates with an inductor and a second capacitor so the other switch can turn off with AC, a resistor discharges the first capacitor.

It turns something simple into something not so simple, but it's an option.

This is what I'm talking about. There it's intended for a high powered circuit breaker, but I don't see why it wouldn't work for a simple bistable mechanical switch. The second capacitor isn't really needed I guess.

[nctnico]

I believe that the highest loads on the power to my house are AC induction motors:
  Refrigerator
  Freezer
  Central air conditioner compressor.
  Forced-air heating fans on gas-fired furnace.
  Vacuum cleaner.
  Clothes washing machine.
  Motor for gas-fired clothes dryer.

What advantage to me is the expensive replacement of all of these to accommodate DC mains?
Transformers may be heavy, but they can be extremely reliable compared to anything using solid-state power electronics.

Most of those appliances use electronics to drive the motor nowadays. Even my 25+ year old washing machine has that.

[Siwastaja]

I believe that the highest loads on the power to my house are AC induction motors:

  Refrigerator
  Freezer
BLDC drives are getting slowly more popular, AC induction motors still very popular
  Central air conditioner compressor.
Nearly all new designs have been BLDC for over 10 years.
  Vacuum cleaner.
I don't think AC induction motors have ever been used for vacuum cleaners. They just need very high RPM, and therefore use brushed universal motors. Work fine with DC; have done exactly that - but power switch will possibly end up in flames so needs to be bypassed.
  Clothes washing machine.
I have seen AC induction motor based clothes washing machine once: an ancient thing with dual-speed windings, and it left the clothes extremely wet because of crappy spin speed. Maybe they are more common in America, but really brushed universal motors are used because that's the only decent way to get the large dynamic range of different speeds, including spin cycle, with small and inexpensive motor.

I'm not suggesting you could just run these appliances with DC supply. Significant modifications will be required nevertheless, even if they used motor types suitable for DC (inverter circuits which just rectify the mains, or brushed series-wound motors which run on DC and turn in the same direction regardless of polarity).

[Gyro]

It might make sense in the far future if all motors move to brushless DC and if semiconductors for working with DC mains voltages at low Rds become cheaper than mechanical switches.
As others have said there are mechanically switching problems with DC so that would all have to go.

Electrolysis is an interesting one that I'd not considered, yeah would have to deal with that somehow.

Electrolytic corrosion is something the London Underground system took very seriously from the start of its post-steam operation. The proximity of so many pipes etc. not to mention the cast iron tunnels themselves means that no traction current (@ 600V DC) is allowed to pass through the running rails. Instead, return current passes through a separate insulated conductor rail - the center one in the attached image. The live rail is the one furthest from the platform. Note the passenger fall refuge pit beneath the center rail (contrary to popular belief, you don't get fried if you touch it on the way down [as it is at nominally ground potential]).

I don't know if other subway systems adopted the same practice.

[David Hess]

There was a coalition of companies that were trying to create a new standard for low voltage DC in the homes like a 30V range USB to power DC products.  There would be one main converter for AC to DC instead of all these wall warts. Probably adaptable to solar too.  Haven't heard anything about that in years.

I suspect any solution which requires more wire runs is uneconomical.

[Zero999]

This has already been discussion.
https://www.eevblog.com/forum/chat/will-ac-power-distribution-become-obsolete/

In short. It's a bad idea and will never happen.

[coppercone2]

it can happen with the right technology. Like an anti electron emissivity switch coating that makes arcing 10x more difficult to bring the switch in line with the AC switch. It would take a break through in materials science to make it possible for the same ease as AC. Otherwise you need some clever design that is more complex.

Don't ask me how or if its possible to make something wanna arc less but also be a good contact. The alternative is some kind of quencher or diffuser for plasma. Kind of like blowing out a flame. Can it be done electromagnetically? Right now the best way is to use some kind of gas, or to make a trap that the plasma goes into and cools down in (parabola shapes).

Tungsten silver is a very tough alloy for switch contacts, but you need to physically deal with the arc.

[TimFox]

I believe that the highest loads on the power to my house are AC induction motors:
  Refrigerator
  Freezer
  Central air conditioner compressor.
  Forced-air heating fans on gas-fired furnace.
  Vacuum cleaner.
  Clothes washing machine.
  Motor for gas-fired clothes dryer.

What advantage to me is the expensive replacement of all of these to accommodate DC mains?
Transformers may be heavy, but they can be extremely reliable compared to anything using solid-state power electronics.

Most of those appliances use electronics to drive the motor nowadays. Even my 25+ year old washing machine has that.

In my case, none of them is rated for DC power input.
“Cold, dead hands”

[bdunham7]

Most of those appliances use electronics to drive the motor nowadays. Even my 25+ year old washing machine has that.

I believe Tim was referring to his own household, not the current offerings in a European appliance store.  In any case, the only one I think he is mistaken on is the vacuum cleaner.

As for the other appliances, although there are BLDC/inverter options for many of them there are still also many of them sold new here in the US that still use PSC or cap-start induction motors.  Air conditioning compressors, top-loading washers that use a transmission, dryers, refrigerators and freezers are all examples and in many cases, the old-school induction motors are only slightly less efficient, much less expensive and much more reliable. IDK exactly why, but that reliability gap seems to be a USA thing.

The one application that seems to have gone mostly to BLDC is the furnace or air handler fan motor that circulates the air in a central HVAC system.  These previously used a multi-speed induction motor and I think BLDC is actually cheaper to make.

In any case, induction motors wouldn't be the only concern.  DC would cause huge issues with other appliances.  Ordinary toasters would go up in flames once your toast was done and a DC-capable toaster would probaby cost 10X what a typical disposable countertop toaster goes for these days.  And then there are microwave ovens...

[Zero999]

There's also an issue with RMS vs peak voltage. Some devices designed to work off rectified 230VAC, might require a higher voltage, than the RMS voltage. A mains powered LED lamp, might have over 100 LEDs in series, each with voltage drop of 3V, so will not work from 230V DC and will require over 300VDC.

[ejeffrey]

There was a coalition of companies that were trying to create a new standard for low voltage DC in the homes like a 30V range USB to power DC products.  There would be one main converter for AC to DC instead of all these wall warts. Probably adaptable to solar too.  Haven't heard anything about that in years.

It doesn't really work.

First off, 30 V is a pretty nice choice.  It definitely has arcing problems with switches but nowhere near as bad as even 48V, much less mains voltage DC. 

However, nothing actually runs on 30V.  So everything still needs one or more step down converters to get to their operations voltages.  These converters can be smaller and cheaper because they don't need as much bulk capacitance.  They probably do need to be isolated to reduce ground loops but they don't need the same level of safety isolation as mains voltage supplies.  You could get rid of wall warts, but the you couldn't use the skinny little power cords most wall warts have, you would need something like AWG16 lamp cord.  Maybe you could use thinner insulation but the insulation is also physical protection against wear leading to short circuits which would be just as much of a problem.

Except 30V is too low of a voltage for many applications.  If we limit ourselves to the standard 15 amp (12 amp continuous) circuit ratings in the US, that is 360 watts per circuit.  You can run small electronics and LED lighting off of that pretty easily.  But it's not enough for something like a desktop computer, kitchen appliances, and so on.  Even a laptop computer with a docking station, a couple external monitors and some other peripherals will tax a 30 V circuit. 

So you either need to have the low voltage DC exist alongside AC mains distributions which costs more, or bump the DC voltage up to something more like 120VDC.  But then you get all the arcing problems in full force, plus all your points of load power supplies now need to be safety isolated.  Which basically means all the wall warts would be back with a vengeance.

There is a way you can do this.  You can get power outlets with built in USB charge ports.  Nobody uses them except at airports because 2 ports is hardly enogh.  But you could make an in-wall low voltage power supply that had 10 USB-C ports, with up to 100 watts per port.  I doubt they would be very popular either.  If cable management is your goal it's better to build that into your desk with a single 120/240V cord going to the wall.

[Siwastaja]

There was a coalition of companies that were trying to create a new standard for low voltage DC in the homes like a 30V range USB to power DC products.  There would be one main converter for AC to DC instead of all these wall warts. Probably adaptable to solar too.  Haven't heard anything about that in years.

It doesn't really work.

First off, 30 V is a pretty nice choice.  It definitely has arcing problems with switches but nowhere near as bad as even 48V, much less mains voltage DC. 

However, nothing actually runs on 30V.  So everything still needs one or more step down converters to get to their operations voltages.  These converters can be smaller and cheaper because they don't need as much bulk capacitance.  They probably do need to be isolated to reduce ground loops but they don't need the same level of safety isolation as mains voltage supplies.  You could get rid of wall warts, but the you couldn't use the skinny little power cords most wall warts have, you would need something like AWG16 lamp cord.

Yeah. It's amazing how a relatively well-engineered, legal and safe 5-10W wallwart from universal 100-250VAC to 5V can be sold to end users for 5USD. Maybe the new fancy 30VDC system would reduce that to 3USD. And only if we do the full switch overnight to create the same economy of scales. Which also won't ever happen because we can't replace 100-240VAC with 30VDC at all; it would come as an additional extra system. All wiring work, sockets, switches, fuses etc. would also come as an extra over the existing AC system which cannot be removed.

And I don't think running the 30VDC system even at 10-15A is practical at all. While amperage defines thermal limits for cables, energy efficiency is important for other (obvious economic and environmental) reasons. And that is relative to voltage squared. Given similar resistance of cables and connectors, 1V drop on 120V is mere 0.8% but 1V of 30V is 3.3%, not insignificant anymore. So such system would probably run at 5A max or something, also allowing nice small connectors. That's just 150W.

But USB-C is quite nice already, allowing us to power phones and laptops of any brand and size from one standard connector, which does the really wise thing and autonegotiate the correct voltage - much more usable for product designers than some forced 30V system.

[Zero999]

The thing is, it wouldn't even save much in terms of losses. The transmission losses might be slightly lower, but the DC:DC converters required to convert 132kV to 11kv to 230V DC etc. would be less efficient, compared to the equivalent transformers.

[Marco]

Wouldn't have had problems with grid following and grid forming and associated stability problems though. At DC, everything would be grid forming.

Geomagnetic storm induced low frequency currents just add or subtract a little power to a DC grid, very little, so those wouldn't have required expensive protection to protect against either.

We're stuck with AC, but it has more problems than just the mismatch with inverter and DC loads.

[TimFox]

DC is often used for long-distance transmission lines for grid-related reasons.
However, switching from AC to DC at the customer end entails huge expense for no real benefit to the consumer.
Other posters insist that the US switch the standard AC voltage at residential and low-voltage commercial service from 120 V to 240 V to make it more European, with a trivial ultimate increase in efficiency, but don't mention possible dangers of the higher voltage.
Funny that no one insists that the other side switch from 50 to 60 Hz, or vice-versa.
In Japan, they have lived with both standards (Tokyo at 50 Hz and Osaka at 60 Hz), which may have given Japanese electrical exporters an edge over mono-frequency countries.