In short
- kW billing charges for real power — the part doing useful work. Power factor does not affect the determinant.
- kVA billing charges for apparent power — the total the system has to carry. Poor power factor inflates the determinant directly.
- The unit is stated in the tariff and printed on the bill.
- Under kVA billing, correction produces a continuous proportional saving. Under kW billing, it produces none at all on that line.
- A capacitor project justified on the wrong unit is the most common avoidable error in this area.
Two utilities can price demand identically in every respect except the unit, and the consequence for a site with poor power factor is a bill difference of tens of thousands a year — and an entirely different answer to whether correction is worth buying.
The difference in one paragraph
Real power, in kilowatts, is the component doing work. Apparent power, in kilovolt-amperes, is what the conductors and transformers actually carry, including the reactive component that does no work. The two are equal only at unity power factor; below it, apparent power is always larger.
A tariff billing in kW measures the work. A tariff billing in kVA measures the occupancy. The second is arguably the better proxy for the cost the utility incurred, because it is apparent power that sizes the equipment.
What it does to the bill
The same load under both units
900 kW of real power at a power factor of 0.80.
- Real power at the peak interval900 kW
- Power factor0.80
- (Apparent power: 900 ÷ 0.80)1,125 kVA
- Schedule billing in kW, rate$16.20 / kW-month
- (Demand charge, kW basis)$14,580
- Schedule billing in kVA, rate$13.90 / kVA-month
- (Demand charge, kVA basis)$15,638
- (Difference per month)$1,058
Additional annual cost of the kVA basis, uncorrected$12,696
Note the kVA rate is lower per unit — that is normal, because it applies to a larger quantity. At unity power factor the two schedules would cost almost the same. The whole difference is the power factor. Figures illustrative.
The last line is the design intent. A kVA tariff is close to neutral for a site with good power factor and progressively more expensive as power factor falls. It is a penalty structure that needs no penalty clause.
Why it decides the capacitor question
Under kVA billing, correction reduces the determinant proportionally and continuously. Improve power factor from 0.80 to 0.95 and apparent power falls from 1,125 kVA to 947 kVA for the same work — a 178 kVA reduction, billed every month, with no threshold to clear and no cliff to fall off. The saving is smooth, predictable and easy to model.
Under kW billing, correction changes the demand determinant not at all. The real power doing the work is exactly what correction leaves alone, by design. A capacitor bank installed at a kW-billed site with no separate penalty clause produces no demand saving whatsoever.
That is the entire decision, and it is settled by reading one line of the tariff. Sites have bought correction equipment on the strength of a proposal that did not check it, which is why the point is labored here: power factor penalties explained sets out the four possible tariff treatments, and sizing a capacitor bank covers what to do once you have confirmed a benefit exists.
The third case
Some tariffs bill in kW and then apply an adjustment when power factor falls below a target — typically multiplying measured kilowatts by the ratio of a target power factor to the actual one.
The effect resembles kVA billing but the arithmetic differs, and so does the optimal correction target. Under a straight kVA tariff, every increment of correction earns. Under an adjustment clause, correction earns until the actual power factor reaches the target stated in the clause, after which the adjustment stops applying and further correction earns nothing.
Reading which of the three you have is a five-minute exercise in the demand section of your rate schedule: how to read a utility tariff book.
Power factor still matters on a kW tariff
Just not on the bill.
Apparent power is what your own transformers, cables and switchgear carry. A site at 0.80 power factor is using 25 percent more of its own capacity than the work requires, and that headroom is what an expansion needs. Correcting power factor can defer a service upgrade, which is a capital saving rather than an operating one and belongs in a different part of the appraisal.
That benefit is real and it should be claimed honestly — as avoided or deferred capital expenditure, with the deferral period stated — rather than dressed up as a demand saving that the tariff does not provide. A finance function that is told a project defers a $180,000 service upgrade by six years will evaluate that perfectly well. The same function, told the project saves on demand charges and then shown a bill where the demand line has not moved, will remember it when the next proposal arrives.
What to check
Look at the demand line on your bill. If the quantity is in kVA, or the rate is quoted per kVA, you are on apparent power. If it is in kW, check separately whether a power factor penalty or adjustment clause exists elsewhere in the schedule, because it will not be on the demand line.
Then, before any correction project is costed, confirm which of the three mechanisms applies and what threshold or target it specifies. Everything else — sizing, switching, harmonic protection — follows from that answer, and none of it is worth doing until the answer is known.
