120V from both legs

That link does not support your claims. Motors create very low levels of harmonics. Filtering these harmonics is NOT generally worth the cost in energy savings. Correcting power factor (displacement power factor) with capacitors is a good idea since it reduces the magnitude of the current flowing to the motor, reducing losses in the entire system.

Harmonics have no effect on displacement power factor. Period. Quite simple actually. Displacement power factor is defined as the phase angle between the voltage and current at the fundamental frequency. By definition, harmonics are not even included.

Displacement power factor excludes harmonics. Poor displacement power factor is caused by inductive, or capacitive, loads. Yes, I have seen facilities with leading power factors. It is just as troubling as lagging, sometimes even worse.

?????

True, utilities are concerned with the power factor of large customers, the displacement power factor. This is corrected using capacitors. Harmonic filtering is sometimes required in such installations, but not for the reasons you state. It is done to avoid resonances that can damage the capacitor bank or other equipment.

They are quite commonly known in the power quality and power metering fields.

BTW, you might be surprised to find the source of much of the information in the link you provided ;-)

Charles Perry P.E.

|

Umm, minor nit. I said that I do not have full knowledge of the operation of the meter's clever magnetic circuit. I said nothing about having "little knowledge." :) I suggest that you read the spiel before dismissing it. The example requires only a very basic level of electrical knowledge to follow. (Which is not to claim that the conclusion is correct. As I said, maybe they do something clever to compensate to a first order. But in the

| I'll second the debunking. The meter knows how | much energy you are using and you get billed accordingly.

But how does it know? To compute power you need both the current and the potential (voltage). The product can then be accumulated/integrated over time to give energy. Every split phase meter I've seen is a 4-terminal device with no connection to the neutral. It can measure the current in each leg and it can measure the potential between the legs, but it has no way to measure the potential between the neutral and either leg. The neutral's potential cannot be assumed to sit half way between those of the legs. An unbalanced load anywhere on the same transformer can drive it closer to either leg.

For the simple case where the load being served by the meter in question is the only load on the transformer it appears to me that you pay for the energy you use on your side of the meter plus 150% of the energy wasted as heat in the neutral on the utility's side of the meter. For more complicated cases where somebody else is pulling the neutral off center you might be charged for less energy than you actually use. Note that we are talking about rather small numbers here. It is not, in any case, something to get excited about.

Dan Lanciani ddl@danlan.*com

Dan the meter doesn't need to measure neutral current. Mr Kirchoff says any current going in will come out somewhere. If it comes out the opposite phase that is a 240v load, any unbalance is a 120v load. You have the current and the voltage. Unbalanced load makes the meter run x speed, balanced load makes it run 2x speed. (or however you want to think about it)

message

following

the

low levels of

worth the cost in

factor) with

the current

on

But

industrial

residential

Period. Quite simple

angle between

definition,

point

there

based on

displacement power

have seen

troubling as lagging,

address

correction

customers, the

capacitors. Harmonic

not for the

damage the

this

net

cost

metering

the information in

message

following

the

low levels of

worth the cost in

factor) with

the current

on

But

industrial

residential

Period. Quite simple

angle between

definition,

I think you might wish to read a little more on the issue.

point

there

based on

displacement power

have seen

troubling as lagging,

To clarify... another term is used commonly 'back EMF' thats a out of phase current that 'travels' back up the feeders to the larger local grid and in many cases to the local area utility grid. From what I can see of your responses so far you are quite bright but just on well read on this narrow spectrum of issues.

address

correction

customers, the

capacitors. Harmonic

not for the

damage the

You are quite knowledgeable..but in my view short on a narrow range of issues in this area.

this

net

cost

metering

correct

the information in

I just scanned the first few links and posted that one as typical... there were 45,000 other hits. My information comes from my experience consulting industrial facilities etc.

I would agree with you that one source seldom has entirely complete and comprehensive data. One should do thier own research. I always have this problem when I post a single link..there are always some nit or other to pick.. so I post the search string also.

Phil Scott

kwh meter used

not 'run the tab

when a very poor

they add

usually an issue

service

charge

local utility.

installed with paybacks

they get a lot of

capacitance in the filters

field current

power

affect other

high source

very minor extent.

cause a small

sequence components

can't confirm

into it for a

loss due to 5th

residences.

it seems that harmonic distortions will have virtually no effect on straight resistance loads...but with motor loads where rotor and stator timing is crucial, harmonic distortions upset the magnetic effects to the rotor, increasing slip... a slight amount of slip changes the back EMF though the stator..resulting in increased current flow.

With compressor motors this can go south fast for various reasons

In my years of looking into these issues I find virtually no EE's that are used to working in this area except for specialists..and there are not too many...you dont see them typically on the usual project. The data is fully out there along with the appropriate test equipment however.

Phil Scott

| | Having read through 'Dan Lanciani's discussion, I can say that what he's | describing is *not* a metering problem. He discusses if the voltage drop | from the meter to the load is significant,

No, I clearly stated that the long run is from the meter to the transformer. The voltage drop that is significant is from the transformer to the meter, not from the meter to the load. You can't change the example circuit and then complain that the math is no longer correct. :)

| True, the losses in the utility equipment are different also, but they are | upstream of the meter and we assume the voltage at the meter is constant.

But that assumption is false. Not only are the voltages not constant but the two line to neutral voltages likely aren't even equal.

| If we don't assume a constant voltage at the meter the numbers change only | slightly, but the results are similar.

Except that the customer pays for some of the utility's line loss, absent some clever correction that has yet to be described.

Dan Lanciani ddl@danlan.*com

|Dan the meter doesn't need to measure neutral current.

I never said it does. What it needs to know is the potential of the neutral relative to a leg. To compute power you need both the current *and* the voltage.

|Mr Kirchoff says any |current going in will come out somewhere. If it comes out the opposite phase |that is a 240v load, any unbalance is a 120v load. You have the current and the |voltage.

No, you don't have the voltage. You are assuming that the potentials between each leg and neutral are equal (much as the meter must). But if there is any current in the neutral, those potentials are not equal. That's where the error comes from. You cannot just take the leg-to-leg potential and divide by 2 to get the leg-to-neutral potential.

Dan Lanciani ddl@danlan.*com

I think you may need to get a copy of IEEE std 100. I will give you a little clue, there is more than one kind of power factor.

Charles Perry P.E.

It would appear that you really mean inductive, or reactive, current. If that is the case, you really should try using industry standard terminology.

From reading all of your posts, I get the impression that you sell some kind of "black box" energy efficiency product.

Charles Perry P.E.

Harmonic distortion in the voltage supplied to a motor will reduce its efficiency and increase heating. If the voltage harmonics are maintained within limits of IEEE STD 519, the effects are negligable.

Charles Perry P.E.

I didn't check Dan's math for errors, but his main premise is correct. Neutral losses on both sides of the meter are measured by the meter in this case. I'm sure that Mr. Perry will agree that the standard split phase house meter violates Blondel's theorem by using 1.5 elements to meter a three wire circuit, or one voltage coil too few. According to the Westinghouse Distribution Systems Reference Book the error due to voltage unbalance is "usually negligible." Where the error is objectionable, a two element meter or two single element meters should be used.

The premise presented in Mr. Doe's original post is incorrect, but if you really want to balance your load, use a 240/120 transformer to serve the

transformer.

Dan,

Just noticed you joined the thread. The old timer in question is not Kirchoff, but Blondel. Might want to Google him. Congrats on your observations.

Here's a little ammo from ANSI C12.1:

A.2.2.3 Single-phase three-wire circuits with balanced voltages A departure from Blondels theorem may be used in a three-wire direct-current or single-phase circuit, provided that the voltages are balanced within acceptable limits. Under these conditions a single-stator three- wire meter may be used, having one voltage coil connected between the two ungrounded wires, and a two- section current winding consisting of two coils wound in opposite directions on a common core. Thus, when each of the current coils is connected in series with each of the line wires of the three-wire circuit, the magnetic effects of the currents flowing in the two coils are additive. The total number of turns of the two coils is the same as would be used on a single-winding two-wire stator of the same current rating. The accuracy of this method is independent of current or power factor balance, but it is dependent on voltage balance.

| >

|

| > Could be (I didn't actually read Dan Lanciani's spiel in your link - | though | > if it makes you more comfortable you can consider that Dan Lanciani | > acknowledges having little knowledge while Mr. Perry and others here are | > very much knowledgeable). I'll second the debunking. The meter knows how | > much energy you are using and you get billed accordingly. Certainly as | far | > as unbalanced load is concerned anyways. A simple way to verify it is to | > just call the utility. I'm sure they'll happily tell you on the phone | that | > it works that way, and they may be able provide some document, or website, | > to that effect. I am assuming you would be content that the utility would | > not outright lie to you on the matter. | >

| > j | >

| | I didn't check Dan's math for errors, but his main premise is correct. | Neutral losses on both sides of the meter are measured by the meter in this | case.

Moreover, I think you pay for 150% of the loss in the utility's neutral because of the error introduced by simplified (i.e., 4 terminal) split-phase metering.

| I'm sure that Mr. Perry will agree that the standard split phase house | meter violates Blondel's theorem by using 1.5 elements to meter a three wire | circuit, or one voltage coil too few.

Exactly. But the nice thing about this problem is that you can show that the meter must be in error without Blondel's theorem or even Kirchoff's law. I tried to word the example so the only thing you really have to take on faith (well, beyond the definition of power) is Ohm's law.

By the way, you can find lots of Blondel resources on the web, including some nice pictures showing that the typical 4-wire split-phase residential meter is not compliant. And various vendor sites point out that non-compilant applications of their metering equipment will not yield true power but have errors "consistent with industry metering standards" in the face of unbalanced loads. Oddly, none of them seem to attach a simple number to the error.

Dan Lanciani ddl@danlan.*com

You need only look at the metering bible "Handbook for Electricity Metering" from EEI. The answer is righ there on page 139 (of the ninth addition). The error will be proportional to 1/2 the difference in the voltages.

Charles Perry P.E.

Okay, I somehow misread that before. You are correct if the voltage imbalance is at the meter (caused by resistance/reactance upstream of the meter), then it won't correctly register such an unbalanced load.

I don't know of any 'trick' in electromechanical kwh meters that can detect this problem. An electronic meter *could* be configured if it senses the current imbalance and has a manual input for the neutral resistance. But I haven't heard of any digital meter algorithm that has this feature.

Of course, two single phase meters, one on each leg to neutral could correctly meter this situation.

daestrom

Such power factor correction works well with motors that have varying load. When not operating at their design point, the pf and efficiency of the motor drops significantly. Such pf correction works by reducing the *real* power drawn from the utility though, not by improving the pf of the current at the meter.

The improvements in billing are *not* a result of the meter incorrectly registering energy consumption. The improvements provided by such devices is through reducing the energy consumed. Consume less energy, and regardless of the pf at the meter, it registers less kwh and billings go down.

Again, the meter will register a higher *real* load because of the drop in motor efficiency. I agree. But a low pf at the meter does not, in and of itself, cause the meter to register differently.

The pf at the meter makes little difference to how well it registers the kwh consumed in the facility. If the consumer wants to find ways to improve their energy usage and consume less energy, *that* will change the kwh registered by the meter. It can be done with a variety of different techniques.

The only flaw in your logic is that the methods you spoke of reduce the energy consumed on the premises and *that* is what reduces the bill. Not changing the pf at the meter (which has *no* effect on the kwh registered/billed).

Pf correction at the load may, or may not cause the load to draw less real power from the utility. But the real power drawn is what registers on the meter. For example, a motor running lightly loaded all the time can have a poor pf. Adding some capacitance across its terminals can improve the pf, but may *not* change the efficiency of the motor. Thus, improving the pf in this fashion does not change energy consumption. In fact the currents into the capacitor bank may add to the total I^2R losses in the system and increase energy consumption.

But pf correction of a motor by solid-state variable voltage control *can* increase the efficiency of the motor at a given load. Such a technique

To repeat, the power factor *at the meter* does not impact how well the meter registers the *real* energy delivered to the premises.

daestrom

The main illogic (other than some numerical problems) is that it is assumed that the meter is somehow measuring power loss "upstream" of the meter. It doesn't- It measures the power on the consumer side of the meter. This will include losses on the consumer's side - for which the consumer is responsible - not those on the utility side. I suggest that you check Charles Perry's reference

Don Kelly snipped-for-privacy@peeshaw.ca remove the urine to answer

This one deals with unbalanced phases- not harmonics- different animals. Certainly unbalance will cause negative sequence currents which are quite undesirable but these are not harmonic currents.