Power factor and kVA billing for business

A ratio measured in two units, and the bill written in the larger one

A commercial meter records two quantities at once. One is the power that actually does work, measured in kilowatts. The other is the larger quantity the wiring and transformers must carry to deliver that work, measured in kilovolt-amperes. The ratio between them is the power factor, and it exists because alternating current loads of the motor-and-transformer kind draw current that is out of step with the voltage, so the current is larger than the work requires. The portion in step produces kilowatts; the out-of-step portion is reactive power, counted in kilovolt-amperes reactive. Divide kilowatts by kilovolt-amperes and the answer runs between zero and one. A site at 0.80 draws 1,250 kilovolt-amperes to do 1,000 kilowatts of work; the same site at 0.95 draws about 1,053. The work is identical and the apparent power is roughly a sixth lower.

• Power factor = kW ÷ kVA, so kVA = kW ÷ power factor.
• Real power, in kW, is what the equipment converts into work, heat or light.
• Apparent power, in kVA, is what the wires, transformers and switchgear must be sized to carry.
• Reactive power, in kVAR, is the out-of-step component; capacitance supplies it locally instead of drawing it through the meter.

Five tariff designs, five different bills for the same building

There is no standard power factor clause. Reading the tariffs side by side shows at least five distinct designs, and they do not produce the same number. Austin Energy scales the billed quantity: where power factor is below 90 per cent during the interval of greatest monthly use, billed kilowatts are the demand in that interval multiplied by 90 per cent and divided by the recorded power factor, and the adjustment reaches the electric delivery charge, the demand charge and the regulatory charge. CPS Energy adds to the demand instead: below 85 per cent lagging the billing demand may be increased by one per cent of actual demand for each point below the threshold. Delmarva Power runs a symmetric adjustment — a per-kilowatt charge or credit for each point away from a 90 per cent base. Duquesne Light bills a billing demand that is the actual demand multiplied by the power factor, so a poor ratio is inside the billed number rather than beside it. Dominion Energy Virginia's large general service schedule can set a minimum distribution demand of 85 per cent of maximum kVA demand where power factor falls below 85 per cent.

• Scale the billed kilowatts by the ratio of threshold to recorded power factor — Austin Energy.
• Add a percentage to billing demand per point below the threshold — CPS Energy at 85 per cent.
• Charge or credit per kilowatt for each point away from a base — Delmarva Power at 90 per cent.
• Bill demand in kVA so the ratio is inside the unit — Duquesne Light's commercial billing demand.
• Set a floor under distribution demand from kVA capacity — Dominion's GS-3 schedule.

Which interval the tariff measures, and why the monthly average misleads

The measurement window matters as much as the formula. Austin Energy applies its adjustment using the power factor during the interval of greatest monthly use, not an average across the month. Oncor's delivery tariff carries the provision at section 5.5.5 and permits an adjustment where power factor is below 95 per cent, calculated from the average power factor occurring during the fifteen-minute peak-kilowatt interval of the monthly billing cycle, applied after ratchets and transformer losses are taken into account. Delmarva measures the average power factor under the operating conditions of the load at the metering point, against a 90 per cent lagging floor, with the adjustment engaging once measured demand reaches 300 kilowatts in the current month or any of the previous eleven. A site can therefore average a respectable ratio across the month and still be penalised, because the penalty is set in the one interval that decides the demand charge.

• Ask which interval the clause names: the monthly peak interval is the usual answer, not the monthly average.
• A ratchet can raise the demand the adjustment is applied to, so the two clauses compound.
• Delmarva's 300 kW trigger means a site can cross into the adjustment as it grows, without any change in equipment.

What actually makes a power factor low

Austin Energy's own explanation names the causes plainly: motors that operate at less than full load are the main contributor, with induction motors, compressors, conveyors, magnetically ballasted fluorescent lighting and various kinds of manufacturing equipment alongside them, and step-down or step-up transformers contributing as well. Delmarva's tariff names lighting, welding and motors in the same breath. The pattern behind that list is partial loading. A motor sized for a peak that rarely arrives spends most of its hours at a fraction of rated load, and an induction motor's ratio falls as its load falls. That is why a building can be fully occupied, busy and still carry a poor ratio: the equipment is working, just not at the load it was chosen for. It also separates two problems that are often conflated — a displaced current, which capacitance can offset, and a distorted current from switch-mode and drive loads, which is a harmonic question with its own limits.

• Motors running well below rated load are the usual cause, not motors running at rated load.
• Compressors, conveyors, welding sets and magnetic ballasts are the loads named in the tariffs.
• Transformers contribute, so a site with several steps of transformation carries more of it.
• Displacement and distortion are different problems with different remedies.

Correction is equipment on your side of the meter

The remedy is capacitance installed at the site, which supplies reactive power locally so less of it travels through the meter. The duty is placed on the customer: Delmarva's tariff requires a customer installing lighting, welding, motors or other low-power-factor equipment to furnish, install and maintain corrective apparatus at its own expense, raising the average ratio to at least 90 per cent lagging. Austin Energy lists the same menu in practical order — resize electric motors, add variable frequency drives, install power factor correction capacitors, replace older magnetic ballasts — and attaches rebate programmes to several of those measures. Where drives and other harmonic sources are present, a plain capacitor bank can interact with the supply impedance, which is why detuned banks with reactors are specified and why harmonic limits at the point of common coupling are set separately, in IEEE 519. Sizing deserves care at the other end too: the tariff language is written for lagging load, and a bank sized to the worst case can run leading when the site is lightly loaded.

• Capacitors supply reactive power at the site; they do not reduce the work the equipment does.
• The tariff puts installation and maintenance on the customer, at the customer's expense.
• Motor resizing and variable frequency drives address the cause; capacitors address the symptom.
• Detuned banks and IEEE 519 limits apply where harmonic-producing loads are present.
• Oversizing risks a leading ratio at light load, which the tariff's lagging language does not contemplate.

Doing the arithmetic on your own interval data

The calculation needs interval data, not a monthly total. Oncor points standard advanced-meter customers to Smart Meter Texas for their interval readings and interval-data-recorder customers to the IDR portal, and most demand-metered accounts have an equivalent route through the utility or the supplier. Pull the fifteen-minute kilowatt and kilovolt-ampere readings for the interval that set the billing demand, divide one by the other, and then apply the tariff's own formula. Austin's arithmetic is a clean example: 900 kilowatts at a ratio of 0.86 gives 900 × 0.90 ÷ 0.86, or about 942 billed kilowatts, while the same 900 kilowatts at 0.92 is billed as 900. Delmarva's tariff adds a procedural point worth knowing before commissioning work — where a demand or power factor is recomputed as the result of an investigation, the new figure applies first to the bill for the month in which the investigation is made.

• Get fifteen-minute kW and kVA (or kVAR) for the interval that set the billing demand.
• Compute the ratio interval by interval; the peak interval is the one that is priced.
• Apply the tariff formula, not a generic one — the designs differ in direction as well as size.
• Ask in writing for the clause number, the interval used, whether a ratchet applied first, and whether a credit exists above the base.

What this changes in a supply decision, and what it does not

Power factor sits on the delivery side of the bill, in the demand quantity and the adjustment clause, and a competitive supplier prices energy and capacity rather than the ratio at your meter. Correcting it therefore does not change the supply rate you are offered. It does change the profile a supplier prices from, which is why the order matters: measure the interval data, correct what the tariff actually penalises, re-measure, and only then take the corrected profile to market. A quote compared against another on a different demand basis is not a comparison, and the demand basis includes this adjustment. Eligibility for competitive supply is a separate question that still depends on the service address, the meter and the utility territory; municipal utilities, cooperatives and some jurisdictions offer no retail choice at all. HighGround Exchange may receive compensation from providers on completed purchases, and an inquiry places no order and switches no supplier.

• The adjustment lives in delivery and demand, so it is not a lever a supplier can price away.
• Correct first, then re-quote: a supplier prices the profile you send.
• Two quotes built on different demand bases are not comparable.

PUT THE OPTIONS SIDE BY SIDE

Your comparison checklist

Common questions