Measuring energy consumption

Some factors here are not quite right.

My understanding is demand is always measured over a period as you say but is real power, not VA as the OP said. If real power, the power factor should not be a major contributor.

"Demand" is a penalty for peaks in the energy use, which requires larger utility generators (amongst other things). Demand can be reduced by temporarily shutting off some loads, like air conditioning cooling, while a required industrial process may have to run at high current for a period.

Utilities don't directly charge a penalty for power factor(which the OP said), but charge for the 'reactive power' (VAR) that is 'used'.

Excellent question in your usual excellent reply.

I had understood, for a mechanical meter, the demand metering was measured from the the rotating power disk on a KWH meter. Is there a separate mechanical VA driver?

Electronic meters would be easy to do either KW or KVA.

My comment was aimed at the OP - that the utility does not actually measure PF.

Most the 'penalty' metering I remember had both a demand scale on the KWH meter and an added VAR meter (KVARH). There is a charge for both demand and VARs. Some commercial metering just had KWH with demand.

I would think utilities might want to measure distortion power factor (harmonic) and add a penalty. The utility can correct for displacement power factor.

-------------- Suppose a daily load is 100KW at 1.0 pf for a 4 hour period of the day or

400KWH and demand of 100KVA Another load has a load of 50KW at 1.0pf but for 8 hours a day. 400KWH and a demand of 50KVA A third case is 50KW at 0.5 pf for 4 hours per day which leads to 200KWH but a demand of 100KVA

Demand is based on peak (over a period ) KVA demand, not KW demand because the sizing of transformers, etc, back to the generator depends on KVA. Thermal limits of equipment are dependent on I^2R heating which is related to KVA at any given voltage (as voltage is kept nearly constant -say +/- 5% normally). Thus both real power and power factor both affect equipment loading and costs (both capital and operating),

----------------------

-------------- Agreed- but ratings of equipment including generators is KVA related rather than KW related (with generators, this is also true- there is a maximum KVA rating but at some point at lower pf leading or lagging, a lower limit exists (lagging- field limits and leading- stability limits - see generator capability curves ) hence load shifting as you indicate to shave peaks, as well as power factor compensation ( beneficial up to the point where the cost of adding compensation matches the savings gained- an exercise in optimization for each specific case).

-------------------------------

Typically, unless things have changed, demand charges (on top of energy charges) are based on KVA peaks rather than reactive alone. Reactive charges don't deal with cases 1 and 2 above and real power demand charges don't deal with the third case. peak KVA demand handles both. However, with modern metering, there could be some benefit to charges for KWH, peak KW and KVARH vs KWH and peak KVA as this would bring home the costs of poor power factor which is controllable independent of KW peaks. There are also associated liabilities involved.

------------ I can see this for KW demand which may now be used for some residential situations. However, I am going to show my age in that the demand meter that I tested in a lab over 50 years ago did measure KVA . As I recall, vaguely, these meters were thermal and took (deliberately) several minutes to reach full scale so that short transients had little or no effect). The meter pointer pushed a "dead" pointer up the scale, leaving it at the maximum reading during the billing period and the meter reader then reset the pointer. As you say, electronic meters are capable of handling this.

I have seen, online, meter which consisted of a KWH meter and a built in thermal demand meter which measured peak KVA independently of KW

------------------------ > Electronic meters would be easy to do either KW or KVA.

------

---------- Certainly the utility can correct for what you call displacement power factor. It can also correct for distortion power factor. Distinguishing between them is another problem. However neither of these are the utilities fault but are due to the customer's equipment and it is easier and fairer to measure KVA demand and put a price on it. The customer can choose to reduce such charges, if displacement power factor is the cause, by putting in corrective equipment, just as, in many cases, local generation can be used to shave peaks. How much compensation is an economic balance. Generally customers are limited as to the harmonic distortion that they cause- and it is also up to them to meet these limits. Excess distortion can have appreciable affects on nearby customers who also produce harmonics- it's a messy interaction. In my back closet I have a copy of a PhD thesis dealing with harmonic analysis of unbalanced power systems written by a consultant who had to deal with these problems in power distribution in industrial areas and returned to school to explore the modeling of such problems.

The utility does use capacitor banks in many situations- for example: distribution radials where voltages may be too low at the far end without them and individual loads do not warrant demand charges.

Interesting link.

I dug out an old "IEEE Recommended Practices..." book (35 years old) and it talks about "power demand" and "energy used over the demand interval". Apparently both KVA and KW demand were used.

Is there a better term to distinguish phase-shift ("displacement") power factor?

I didn't know utilities were correcting distortion power factor. How do they do it?

The European Union, I believe, has strict limits on equipment power factor, which limits harmonic distortion. I would think utilities would be big promoters of limits on manufacturers this side the pond.

All you are talking about is raising the PF to lower the current to lower the voltage drop? I guess I think of caps to lower current (which of course lowers voltage drop). I don't think of them as voltage-raising devices. (I guess I should.)

================ If a rotating disc VARH meter has capacitive load I believe it runs backwards. Do you know if utilities allow that? (It could be blocked mechanically.) Are electronic VARH meters allowed to run backwards?

I never really got into billing/interconnection metering but this I do know. Mechanical meters can be fitted with a device (detent, IIRC) that prevents 'backward' rotation. Backward meaning counting in the 'wrong' direction when power flow is reversed.

A pumped storage plant that I helped design and start up while I was still in college, had a minor crisis when it was discovered late in the design that an additional 4 ft panel would be required to hold all the WH meters, test switches, pulse counters and transformer/line loss compensators for separate metering and telemetering, in and out, of each generator/pump.

---------

-------------- The same way that individual plants do it- use filters. However I believe that in general the onus for this is placed on the plant causing problems.

--------------------

---------- I would suggest that these limits are on harmonic content rather than on power factor and such limits are put on customers who have harmonic generating equipment- they can have harmonics as long as the harmonic level "seen" by the utility is within limits. Balanced harmonics are considered in IEEE Standard 519 (as of 1995) and "residual harmonics" due to unbalance effects are considered in some jurisdictions which have tougher standards than this. This of course has nothing to do with conventional pf (displacement) which leads to other problems. I would suggest that pf is not a good measure (e.g. an induction motor may go from 50% to 90 % pf from no load to full load).

---------------.

---------- I think of it as shunt capacitors supplying all or some of the inductive reactive required by loads or I^2X in lines beyond the capacitor location. Yes- this reduces voltage drop and can in some situations result in a leading pf which raises voltage. Effectively they supply reactive locally rather than having it supplied from a more remote location.

A question in my mind is that I really don't see a great deal of use in measuring VARH but if such measurement is considered useful , why not allow reversal? VARH doesn't measure anything with respect to equipment losses and sizing. WH measurement is a measurement of energy delivered which is the whole purpose of the electrical grid. KVAR and KW measurement is useful. Pf measurement does give useful information but typically it is derived from two of KW, KVAR and KVA. There have been pf meters around for a long time but typically, those that I have seen are not useful for anything but indication. With electronic metering, since power and reactive measurements can be made digitally - pf is a simple calculation. Note that with 3 phase systems, total power and Vars can be measured accurately even when the system is unbalanced. However, in the unbalanced case, there is no single pf as pf is essentially a single phase concept.

Understandable if "beyond" is upstream. If "beyond" is downstream from the caps, how is the current changed downstream?

That is what I don't understand. Why does leading PF raise voltage? Why doesn't leading PF increase the current over PF=1, which produces more voltage drop?

I assume we are talking about the same thing - 'reactive power' - in effect a KWH meter with the voltage to the meter at 90 degrees. The "VA" is deceptive.

I believe VARH measurement is fairly common here for industrial, and is used to charge a significant penalty to encourage PF correction within the plant.

Utilities might want a higher penalty for VAR 'use' than credit for, in effect, the plant correcting the utility PF. It is somewhat like if the plant had its own generator with excess capacity and supplied power back to the utility, the utility would not want to buy the power (by KWH meter reversal) at the same rate as the plant bought energy. But in the case of VARH it is not obvious to me what the utility would want to do, hence the question.

I believe overcorrection can cause resonances at harmonic frequencies, so PF correction control equipment probably avoids overcorrection.

VARH (KVARH) is some measures the amount of PF correction the utility may have to supply (but not the max amount). It is good for charging a penalty to encourage PF correction.

I haven't tried to verify this by calculation but, I suspect that the correct value of capacitance, even tho it under-corrects pf, can cause oscillations. Particularly if there is some harmonic excitation source like an ASD.

-----

------------------ Using capacitors to correct pf is essentially using parallel resonance so ideally such resonance occurs- but the Q is low . This is the desired result- unity pf. At harmonic frequencies the net impedance of load and capacitor will become capacitive (again with a healthy resistive component) and detuning occurs- no resonance. The danger that you appear to indicate would be the situation of a series resonance between line reactance and the capacitive load and this could happen. Would it be a serious problem- depends on the situation- there is no cut and dried, "one size fits all" answer. There would be some increase in magnitude of the harmonics near this resonant frequency but the dominant resistive part of the load would provide heavy de-tuning. The worst case is the situation where the capacitors are on the end of the line with the load disconnected-but then you wouldn't have a harmonic source due to the load site so the problem isn't there. There would, however, be a rise in the fundamental voltage. Note that I don't consider harmonics as "oscillations" in any different sense that the fundamental is also an oscillation. Perhaps you can clarify what you meant.

Don,

You have the correct take on what I'm thinking about, i.e., series resonance. It's been 30 years since I even thought about it and I couldn't remember the right terms.

By 'oscillations' I mean an amplification, by series resonance, of a harmonic voltage. I couldn't think of the proper wording at the time. I suppose that parallel resonance could also be a problem but , having never designed a capacitor bank, I don't know if a normal application would be subject to this.

I have read of ASD harmonics (the excitation source) at one site causing problems in equipment at other sites sometimes many miles away. Poorly designed cap banks can contribute by making inadvertent tuned circuits. I know that series cap banks (not used for pf correction) have caused problems with subsynchronous resonance.

The thing is that the application of series capacitors is only done on long, HV transmission lines and only when it is necessary- such correction is not normally near resonance but even then protection is applied to bypass or short the capacitors when, due to some fault, there is a resonance condition. The sub-synchronous problem that you have referred to is of concern in some situations. Hermann Dommel etc have referred to this electromechanical interaction.