AC Efficiency: Fact or Fiction?

[nsaspook]

as for the video , the first thing I thought was Oh Snap! when they used the 310v DC or thereabout (rectified AC) and the contact was open the switch was like 1cm or more apart and the arc still formed , how is that possible? isn't that arc a bit big for 310v?
Or does DC behave differently in this manner as it's capable of striking longer arcs at the same voltage level?

Sustained electric arcs are non-linear, increased current results in a lower voltage so you have a unstable negative resistance circuit with high current density and low voltage drop across the arc that functions as a positive feedback circuit. With DC (no AC crossing and its always arcing from one cathode hot spot) If drawn apart slowly the arc will maintain its high temperature longer and continue to generate emission at the arc point until it becomes too small.

 

[jim hardy]

Since we're digressing;
here's a piece of trivia for the back of your mind.
If you've ever electric welded you have an intuitive "feel" for arcing.

An arc in an AC circuit will likely go out at next current zero crossing. That's why home 'buzz box' welders are so tricky to get started.
An arc in a DC circuit does not have the benefit of a natural zero crossing. So current must be "brute forced" down to zero . That's why a good industrial DC welder is such a pleasure to use.

That's also why the fastest of AC circuit breakers are rated to interrupt current in ~10 milliseconds - a zero crossing is sure to come by within a half cycle. And that's why the DC rating is so much lower than the AC rating, small breakers like in your household panel are not brutish enough for DC service. Hence Mr spook's point.

There exist especially fast fuses for semiconductor protection. We used Chase Shawmut form 101. They're special shaped silver links in a stout epoxy-glass fiber tube filled with sand. The sand breaks up the arc and melts absorbing the heat , that's called "quenching the arc" .. They'll interrupt current in a millisecond or two.
http://www.ferrazshawmutsales.com/pdfs/A100P.pdf

In 1973 i spent a couple weeks applying short circuits to various protective devices and recording the waveforms.

Ahhh, nostalgia... we had inverters in the plant and were looking for a circuit breaker fast enough that it'd interrupt a short on inverter output before the inverter's internal protection shut it down, which trips the plant. It's embarrassing when somebody innocemtly changing a lightbulb can trip a nuke plant.

Here's a typical fast circuit breaker response

That let-through energy under the gray peak probably wouldn't hurt a motor but it might wreck the semiconductors in an electronic speed controller for that motor.

So you'd use a faster device like that Amptrap fuse , it gives a waveform like the blue segment . Observe how much more gentle that is on the load than the circuit breaker would be. Someplace in my barn is a notebook with my old 'scope photos, but these reproduction traces came from http://www.galco.com/comp/prod/fuses.htm
and are faithful representations of real ones..

999 out of a thousand people will never need to know about this little factoid
but i hope it helps somebody.someday.
Sorry it's not more academic - just a qualitative introduction not quantitative analysis.

old jim

 

[Jeff Rosenbury]

DC power is used in a few cases of long distance power transmission. The cost of motor generator pairs to reconvert it to AC must be less than the additional cost of the extra wire (s?) for three phase power.

Of course my dream is to have every device store it's own power in a super-capacitor. Whenever the capacitor winds down, we teleport it to the recharge station and back in under a µS. This would of course happen many times a second. (Beam me up Scotty!).

There would be predictive software so only the equipment I wanted would show up. I never need to leave my couch again.

Never underestimate the power of the Dark Side -- of my couch.

 

[sophiecentaur]

The cost of motor generator pairs to reconvert it to AC

I thought they use 'electronic means' with mercury arc valves and inductors. (?) @Jim - tell us what's the practice these days.

 

[Jeff Rosenbury]

I thought they use 'electronic means' with mercury arc valves and inductors. (?) @Jim - tell us what's the practice these days.

It varies. You are right that it's been nearly a hundred years since they installed a new motor generator pair in a major installation. Mercury was used for a while, then thyristors. Offshore wind generators are likely to use other (proprietary) means that combine various electronic solutions. HV DC seems to be more common as power inverters drop in price and offshore cables cost more.

 

[jim hardy]

I was never around HVDC

Thyristors have become stout enough for that service

here's a pretty good introduction by a major player in the utility scale equipment field.

http://www.energy.siemens.com/us/pool/hq/power-transmission/HVDC/HVDC_Proven_Technology.pdf

here's page 15 showing the thyristor valves. They give a good description of a British project

There's also the long term savings in energy loss
and the ability to interconnect 60hz with 50 hz systems(as they do in Japan)

One gets to like this seemingly boring field, power
i guess guys are just attracted to big machinery.

I know some railroad enthusiasts, too.

old jim

 

[anorlunda]

We're all over the map. Sophie diverted to LVDC distribution, now we diverted again to HVDC transmission. About as different as possible.

HVDC had advantages including stability, lack of VAR generation, ability to link dissimilar frequencies, and lower losses. Offsetting these is the cost of the converter. One needs a converter at each end of the HVDC line, plus an additional converter at each intermediate point where you want to tap power.

Think of an attractive HVDC line from Idaho to Southern California. All the states in between the end points would receive zero benefit unless they pay for additional converters within their borders. That makes it really really hard to get approval from all those governments. Midwest USA wind power owners advocate for big HVDC lines to New England so that they could sell their green energy there. But the westerners want the easterners to pay for the HVDC. The easterners say, "Heck no. You deliver it to my doorstep if you want to market it" The in-between states crossed by the lines say, "What benefit do we get?" (Sophie, a European parallel would be if Norway wanted to sell power to Italy with HVDC lines crossing Scotland and England but with no electrical connections inside UK borders.)

To calibrate costs see https://en.wikipedia.org/wiki/High-voltage_direct_current

For an 8 GW 40 km link laid under the English Channel, the following are approximate primary equipment costs for a 2000 MW 500 kV bipolar conventional HVDC link (exclude way-leaving, on-shore reinforcement works, consenting, engineering, insurance, etc.)
Converter stations ~£110M (~€120M or $173.7M)

That comes to $86/kw of capacity for the converter. (100x more than $0.80/kw for new solar panels.) That's pretty expensive, so HVDC is used only in cases where its advantages overcome the costs.

 

[sophiecentaur]

Sophie, a European parallel would be if Norway wanted to sell power to Italy with HVDC lines crossing Scotland and England but with no electrical connections inside UK borders.)

I can see that and it is very relevant in some parts but the EU can be surprisingly co operative amongst themselves and I was assuming that any HVDC network would include all countries within.

That comes to $86/kw of capacity for the converter. (100x more than $0.80/kw for new solar panels.) That's pretty expensive, so HVDC is used only in cases where its advantages overcome the costs.

That's a good bit of info. My enthusiasm is a bit dampened - or perhaps delayed by a few extra decades.

 

[arationofreason]

Altho the main argument lies in the difficulty of cheap conversion of DC voltages, remember also that balanced 3 phase AC systems also reduce the amount of transmission conductor required by 25 %.

 

[sophiecentaur]

remember also that balanced 3 phase AC systems also reduce the amount of transmission conductor required by 25 %.

That, no doubt, is a well known bit of book work but it is counter intuitive, Doesn't I²R apply? Or are you just saying that three wires are used instead of four? In which case, it wouldn't be comparing like with like.

 

[russ_watters]

That, no doubt, is a well known bit of book work but it is counter intuitive, Doesn't I²R apply? Or are you just saying that three wires are used instead of four? In which case, it wouldn't be comparing like with like.

Yes, that and the fact that three phase power is volts * amps * √3 means you can use much smaller wires and get the same resistance.

 

[sophiecentaur]

Yes, that and the fact that three phase power is volts * amps * √3 means you can use much smaller wires and get the same resistance.

The root three factor only tells the RMS value, surely (if V and I are in phase). That value is the DC value and the DC dissipation through supply cables will surely be the same as the equivalent AC losses, for a given total load. I appreciate that it is easier to divvy up the loads carried by the three conductors in 3 Phase systems when changing to DC but I think it is possible (not straightforward) to use switch mode techniques to regulate the charge supplied by the same three conductors so that the power is carried evenly. The invertors still have inductive isolation from input to ouput so you can still make use of the potentials between each pair of the three conductors. An expensive exercise, of course (they have all told me that I'm far too optimistic) but I think the 'theory' is there.

 

[russ_watters]

The root three factor only tells the RMS value...

No, the root three factor is due to the fact that the phases are 120 degrees out of phase with each other: the voltage and amperage are already/always measured as RMS.

Single phase power is VI and three phase power is √3*VI. That difference means even with 3 wires instead of 2, you can use less total copper in the three phase system for the same resistance.

 

[Jeff Rosenbury]

DC has less I²R losses for a given amount of metal because there are no complex effects like reactive power or skin effect causing extra losses.

I thought RMS power was the root mean square power of the sine wave. In electronics, we sometimes use other waves and need to figure power differently (When we bother; I've never built a power amp that didn't use sine waves. )

Also one thicker cable costs less than 12 thinner cables. (Each phase has multiple conductors to increase the geometric mean radius, which affects the capacitance, which in turn matters more in AC than DC.)

Also, the Earth can sometimes be used as a ground return path -- I think. (I would look long and hard before I did this, but with hundreds of millions of dollars at stake ... ) This leaves one cable rather than two.

Long lines are better as DC. Short lines are better as AC. Which is better in any given case is a complex decision based on economics and sometimes, as anorlunda wrote, on politics.

 

[sophiecentaur]

No, the root three factor is due to the fact that the phases are 120 degrees out of phase with each other: the voltage and amperage are already/always measured as RMS.

Single phase power is VI and three phase power is √3*VI. That difference means even with 3 wires instead of 2, you can use less total copper in the three phase system for the same resistance.

Could you give me a reference to this (On line)? Trying to work it our for myself got me in a muddle.

 

[anorlunda]

Could you give me a reference to this (On line)? Trying to work it our for myself got me in a muddle.

There is a simpler way to think about it. Transmission costs are more heavily influenced by conductor size (thick wire costs more per km but it reduces $I^2R$ power losses) than by conductor spacing and insulators (whose costs rise with voltage). Therefore, the simpler way is to ignore voltage.

A single-phase circuit 100 km long needs 200 km of wire to carry 1 per-unit power. A three-phase 100 kM circuit using the same conductor size, needs 300 km of wire to carry 3 per unit power. Power capacity is increased x3 while the km of wire is increased x1.5. That is the central advantage of three-phase power.

Note: the same logic could lead us to use more than three phases, and indeed that idea sounds attractive. But there are other costs and complexities, so three phase is the nearly universal winner in the trade-offs.

 

[russ_watters]

Could you give me a reference to this (On line)? Trying to work it our for myself got me in a muddle.

http://www.engineeringtoolbox.com/three-phase-electrical-d_888.html
https://books.google.com/books?id=PsC0bSrj8C8C&pg=PA30&lpg=PA30&dq=three+phase+power+less+copper+than+single&source=bl&ots=4qXWqsGxdb&sig=3vz1vN_66RV6bdAFrYy1VXnPUXQ&hl=en&sa=X&ved=0ahUKEwjWiriZt4HKAhXDGR4KHfQoCAEQ6AEIRzAH#v=onepage&q=three phase power less copper than single&f=false

Note that in the US, residential 240V service is "split phase", with two hot wires at 180 out of phase with each other. This is an even more efficient use of wires as you've doubled the power you can deliver with the same number of wires (vs normal single phase).

 

[anorlunda]

Note that in the US, residential 240V service is "split phase", with two hot wires at 180 out of phase with each other. This is an even more efficient use of wires as you've doubled the power you can deliver with the same number of wires (vs normal single phase).

No no no. There are three wires in that system not two. Look at this diagram below from https://www.physicsforums.com/threads/total-amperage-in-a-service-panel.705961/ post #2 by Drakkith. Now suppose that nothing is plugged into the 240 volt plug. You see that the 120 volt current coming in from hot#2 must return through the neutral, not hot#1. The neutral is not the same as the third ground wire in plugs that is used for safety only.

 

[jim hardy]

@sophiecentaur

Trying to work it our for myself got me in a muddle.

I must've been lucky in 1965. Our grad student AC power instructor spent twenty minutes at the blackboard drawing phasors and explaining it to us boys.
It has to be presented right the first time so it'll be intuitive thereafter and he did a great job.the key to it is rigor in naming, a la Lavoisier.
We must distinguish between
current in an individual phase winding
and
current in an individual line ,
call them I_phase and I_line
also between voltage across a phase winding V_phase and voltage between two lines V_line

when we do that we see
for delta connection
V_line = V_phase
and
I_line = sum of two I_phase 's.plain trigonometry and single step thinking will get I_line to I_phase relationship...

take a delta connected machine, motor or transformer - in any of the three single windings power is VI cosθ and for simplicity assume unity pf (resistor bank?) .
To be more specific in terminology power in each phase is V_phase I_phase.
Total power is 3X V_phase I_phase.

Now what about the wires carrying power to(or from) the device? That's where we'd hook up measuring instruments.
Each wire carries the current for two phase windings.
Call that the line current .
Phase currents are 120 deg out of phase with one another
and if you add two equal phasors head to tail at 120 deg, their sum is √3 not twice their individual magnitudes.

SO I_line is √3I_phase
AHA !
With the delta connection ,
Current in the Line is greater than current in the phases by that ubiquitous √3 !
I_line = I_phase X √3
and I_phase = I_line/√3

So - were i to read ammeters connected in series with the lines
and voltmeters connected between the lines ,
and multiply those two numbers,
wth delta connection i'll get a result
V_phase X (I_phase X √3) because I_line >I_phase by √3
which is neither total power nor power in a single phase

Total power is
V_phase X I_phase X 3,
Since with delta connection, V_phase = V_line
we can write for delta connection
total power = V_line X I_phase X 3...

and since with delta connection I_phase = Iline/√3
we can write for delta connection
total power = Vline X I_line/√3 X 3 = Vline X I_line X√3
plod through it once drawing those phasors and it's intuitive ever after.

Wye connection?
Since for wye connection
I_phase and I_line are equal
it's V_phase and V_line that differ by √3
step by step plodding will get to the exact same expression

KVA3phase = V_line I_line√3

That it's not intuitive to everybody suggests to me that our grad student Charlie Gross was an exceptional teacher. He's at Auburn now and has written several textbooks.

Sorry Sophie I'm too awkward with latex and graphics to draw a picture
the key is that √3 line to phase ratio for either voltage or current depending on Δ-Y

old jim

 

[russ_watters]

No no no. There are three wires in that system not two. Look at this diagram below from https://www.physicsforums.com/threads/total-amperage-in-a-service-panel.705961/ post #2 by Drakkith. Now suppose that nothing is plugged into the 240 volt plug. You see that the 120 volt current coming in from hot#2 must return through the neutral, not hot#1. The neutral is not the same as the third ground wire in plugs that is used for safety only.

Huh? If there is nothing plugged-in to the receptacle, then there is no current flowing at all. My suspicion is that some 240V receptacles use a neutral because for whatever reason they need to be capable of supplying both 120V and 240V. But they don't necessarily need to be: the alternate voltage of a 3-phase system (such as 240V) also uses two hot wires and no neutral. See:

All NEMA 6 devices are three-wire grounding devices (hot-hot-ground) used for 208 V and 240 V circuits...
NEMA 6 devices, while specified as 250 V, may be used for either 208 V or 240 V circuits, generally depending on whether the building has a three-phase or split-phase power supply, respectively.

https://en.wikipedia.org/wiki/NEMA_connector#NEMA_6

 

[Averagesupernova]

I think the point russ is making is that you can have a pair of wires carrying 20 amps at 120 volts supplying 2400 watts to the load. Switch over to the split phase system and you add only one wire to double the power available. The alternative of course is to increase the voltage which is not practical or replace the wires which may not be practical.

 

[russ_watters]

I think the point russ is making...

My post #55 was poorly written at first...I've re-written it now. Should be clearer.

 

[anorlunda]

Huh? If there is nothing plugged-in to the receptacle, then there is no current flowing at all.

No, look at the diagram.. If there are loads plugged into the 120V receptacles, but none in the 240V receptacle, where does the current flow?

The 120V plugs have three prongs, hot#2, neutral, and ground. The ground is for safety. The neutral is for the return path of current.

 

[jim hardy]

I've run across clothes dryers with 240 volt heating element and 120 volt motor

they must have a neutral to return the motor current. That means a 4 wire plug.

 

[russ_watters]

No, look at the diagram.. If there are loads plugged into the 120V receptacles, but none in the 240V receptacle, where does the current flow?

120V power flows from hot to neutral through 120V plugs in 120V circuits. We're not talking about 120V circuits, we're talking about 240V circuits.

The 120V plugs have three prongs, hot#2, neutral, and ground. The ground is for safety. The neutral is for the return path of current.

No, 120V plugs have one hot and one neutral, for a total of two prongs (we've been ignoring the grounds, which some have and some don't). Some 240V plugs have a neutral, some don't. Have another look at my post - I've re-written it and provided an example that makes it clearer.

 

[anorlunda]

Switch over to the split phase system and you add only one wire to double the power available.

I agree, but he misspoke. He said the same number of wires as single-phase, but as you point out we must add a third wire. A three-phase system delivers 3x power with 3 wires. The split system delivers 2x power with 3 wires.

 

[russ_watters]

I've run across clothes dryers with 240 volt heating element and 120 volt motor.

Thanks, that's what I was guessing -- the chart on the wiki page lists clothes driers as having 4-prong plugs (2 hot, neutral, ground). I figured it was so they could provide both 120 and 240V.

 

[russ_watters]

I agree, but he misspoke. He said the same number of wires as single-phase, but as you point out we must add a third wire. A three-phase system delivers 3x power with 3 wires. The split system delivers 2x power with 3 wires.

No, averagesupernova was not correct about what I meant (in fairness, the first writing of the post was not very clear) and no, you don't have to ("must") add a 3rd wire to get 240V. Indeed: you only "must" have a 3rd wire if you want to supply two voltages instead of just the 240V.

 

[jim hardy]

Thanks, that's what I was guessing -- the chart on the wiki page lists clothes driers as having 4-prong plugs (2 hot, neutral, ground).

that's a relatively new code requirement for buildings, so they can accommodate such dryers.
It's handy of you're using the dryer outlet with a big extension cord for a welder or aircompressor - you can get both voltages where you're working,.
One should have a GFCI in such an extension cord
One should never use the green wire to return normal operating current.

 

[Averagesupernova]

Ok this has gotten pretty nutty. About as basic stuff as I can think of and people who I have had a fair amount of confidence in are in some kind of disagreement/misunderstanding or whatever. :(

 

[sophiecentaur]

Note that in the US, residential 240V service is "split phase", with two hot wires at 180 out of phase with each other. This is an even more efficient use of wires as you've doubled the power you can deliver with the same number of wires (vs normal single phase).

This is pretty obvious to me but one has to compare apples with apples. You have doubled the voltage so the available power is double, for the same current. The centre tap doesn't have to play any part in that arrangement and I think it's a red herring. The US system really does seem to bring in another layer of difficulty for people to get their heads round, with a range of strange conclusions about the consequences of such a system.
But I think there could be a similar 'misdirection' in claiming an advantage for three phase and single phase - because the effective volts are different for the two systems. The voltage limit is less obvious than the Current situation because the relevant I is RMS but the relevant Voltage is Peak. I realize that insulation can be a lot cheaper than copper to install.
I will now read those helpful links and try to get myself sorted.

@Old Jim: Thanks for your description. It is getting me there! I will search for a suitable version of the argument with some diagrams and Maths.

 

[sophiecentaur]

Ok this has gotten pretty nutty.

It could appear so but I think it is such a multi-faceted business, with some facets carrying more weight than others, that it's not surprising such a wide open thread title has led us all over the county.

 

[Averagesupernova]

This is pretty obvious to me but one has to compare apples with apples. You have doubled the voltage so the available power is double, for the same current. The centre tap doesn't have to play any part in that arrangement and I think it's a red herring. The US system really does seem to bring in another layer of difficulty for people to get their heads round, with a range of strange conclusions about the consequences of such a system.
But I think there could be a similar 'misdirection' in claiming an advantage for three phase and single phase - because the effective volts are different for the two systems. The voltage limit is less obvious than the Current situation because the relevant I is RMS but the relevant Voltage is Peak. I realize that insulation can be a lot cheaper than copper to install.
I will now read those helpful links and try to get myself sorted.

Huh? Voltage is spec'd as peak in 3 phase but not single or split-phase? Since when?
-
Without the center tap we are no longer maintaining the original voltage as we do in split-phase. It is quite relevant. Also, in a 3 phase wye system there is a center tap and I think it is quite relevant as well.

 

[sophiecentaur]

Huh? Voltage is spec'd as peak in 3 phase but not single or split-phase? Since when?

It is the peak voltage that determines whether or not the insulation will fail. That's what I meant. Insulation for an AC 120V system need not be as good as for an AC 240V (or the split phase that the US use) system. That is a hidden cost for the 'advantage' of doubling the Power capacity in this case.
The three phase 'advantage' is harder to calculate and I am scouring the info that's been given on this thread. Bare assertions won't help my understanding of this.

 

[russ_watters]

Ok this has gotten pretty nutty. About as basic stuff as I can think of and people who I have had a fair amount of confidence in are in some kind of disagreement/misunderstanding or whatever. :(

If you haven't re-read my post #55 since I edited it, please do. It should clarify this. Perhaps people are used to doing split phase 240V with 3 wires (plus ground) and aren't aware it can be done with 2, just like 208V single phase is done with 2 legs of a 3 phase circuit, with no neutral. Otherwise, I think we're probably just talking past each other.

The basic issue we are discussing is whether a 240V split phase system must have a neutral. It can have a neutral if you also want it to provide 120V power, but it doesn't need to have a neutral. Just like if you are using 2 phases of a 208V three-phase circuit to get 208V single phase.