How Utilities Calculate kWh: A Complete Breakdown of Every Input on Your Bill

Every industrial electricity bill is the product of dozens of measurements, calculations and contractual mechanisms, most of them invisible to the person paying it. This report breaks down how a utility arrives at the kWh and kW figures on your invoice: which inputs come straight from the meter, which are calculated under IEEE and IEC standards, which come from the tariff, and how long a power quality improvement takes to reach your bottom line.

01The fundamental measurement: how meters calculate kWh

What the meter actually measures

At its core, a revenue-grade electricity meter measures only two physical quantities directly:

1. Instantaneous voltage, V, sampled on each phase through a voltage transformer (VT, or PT) or a direct connection.
2. Instantaneous current, I, sampled on each phase through a current transformer (CT).

Everything else, kW, kVA, kvar, power factor, THD and kWh, is calculated from these two raw inputs.

Modern digital meters use high-speed analogue-to-digital converters to sample the voltage and current waveforms many times in every cycle, thousands of times a second. That captures the full shape of each waveform, harmonic distortion included.

The kWh calculation

kWh is the time integral of real power. The meter computes instantaneous power, p(t) = v(t) × i(t), at each sample, averages it over each cycle to give real power in kW, and accumulates the result over time:

kWh = ∫₀ᵀ P(t) dt

Here P(t) is the real power at time t, and T is the billing period. The meter performs this integration continuously, adding to the kWh register in real time. Because it samples the whole waveform, distortion included, the kWh figure captures all the real power delivered, fundamental and harmonic.

The pointkWh is a direct, continuous measurement. It reflects the energy consumed in real time, with no averaging window, no delay and no ratchet.

If losses fall at 2:00 pm, the meter registers fewer kWh from 2:00 pm onward. That is fundamentally different from kW demand, as the next section shows.

02kW demand: the running average and why it matters

How kW demand is measured

kW demand is an average over a fixed interval. The meter divides time into demand intervals and calculates the average kW drawn in each one. The highest average recorded in a billing period becomes the maximum demand, the figure the demand charge is billed on.

Standard demand intervals

The 15-minute interval is the most widely used, and the norm in North America, where each utility’s tariff sets it. In Great Britain, the settlement period is half an hour.

Block interval or sliding window

Meters compute demand in one of two ways.

With block metering, a 10-minute spike that straddles two 15-minute blocks is split across both and looks smaller. A sliding window catches it whole. For the same load, sliding-window demand is never lower than block demand, and usually higher, which means a higher demand charge.

Worked exampleA site running at 400 kW takes a 10-minute spike to 1,000 kW, from 00:10 to 00:20. Block metering splits it: each block averages (10 × 400 + 5 × 1,000) ÷ 15 = 600 kW. A 15-minute window sliding in 5-minute steps catches 00:05 to 00:20 whole: (5 × 400 + 10 × 1,000) ÷ 15 = 800 kW.

03Power factor: how it inflates your bill

What the meter calculates

The meter calculates power factor from its voltage and current measurements:

Power factor is not a separate sensor. The meter calculates it from voltage and current, following the definitions in IEEE 1459, which covers sinusoidal, non-sinusoidal, balanced and unbalanced conditions.

How power factor affects billing

Utilities penalise poor power factor through one or more of four mechanisms:

Power factor is calculated from measured data, not measured directly. The threshold at which penalties apply, 0.90 or 0.95 for example, is a contractual parameter from the utility’s rate schedule, not a measurement. IEEE 1459 defines how the power factor itself should be calculated, especially when harmonics are present.

04Harmonic distortion: the hidden energy tax

What the meter measures

The meter computes total harmonic distortion (THD) by running a fast Fourier transform (FFT) on the sampled voltage and current, splitting each waveform into its fundamental and its harmonics. The standard definition, used by IEEE 519-2022, is:

THD = √(Σ I_h²) ÷ I₁ × 100%, where I_h is the RMS current of the h-th harmonic and I₁ that of the fundamental. Harmonics up to the 50th order are usually included, following IEEE 519 and IEC 61000-4-7.

THD is calculated by the meter from its samples. The limits that define acceptable distortion come from IEEE 519-2022, the standard for harmonic control in electric power systems.

How harmonics affect kWh

Harmonics raise kWh through several mechanisms, all of them real losses the meter records:

1. More I²R loss in conductors. Harmonic currents flow in the same conductors as the fundamental and add to the total RMS current, and resistive loss rises with its square. The meter sees it as real power consumed.
2. Eddy-current loss in transformer windings. It rises with the square of the harmonic order: per unit of current, the 5th harmonic, at 250 Hz on a 50 Hz system, causes 25 times the eddy-current loss of the fundamental (IEEE C57.110).
3. Skin effect. At higher frequencies, current crowds towards the surface of a conductor, raising its effective resistance, most of all in large conductors.
4. Neutral loading. Triplen harmonics, the 3rd, 9th, 15th and so on, add together in the neutral instead of cancelling, and can push the neutral current above the phase current.

IEEE 519 limits apply at the point of common coupling (PCC), usually the metering point between the utility and the customer. They are limits set by the standard, not meter readings: the meter measures the distortion, and the standard defines what is acceptable.

05Voltage imbalance between phases

What the meter measures

The meter measures the voltage on each of the three phases independently. Voltage imbalance, or unbalance, is then calculated in one of two standard ways:

The voltage on each phase is measured directly, and the imbalance is calculated from it. The IEC method, symmetrical components, is the more rigorous, and the one IEC 61000-4-30 specifies for Class A power quality instruments.

How voltage imbalance affects your bill

Voltage imbalance rarely appears as a line on the bill, but its effects are captured in the kWh and kW readings:

1. Negative-sequence current in motors. A small voltage imbalance drives a current imbalance several times larger, and the extra current raises the I²R loss in the windings. That loss is real energy, consumed and metered.
2. Motor derating. NEMA MG 1 says a motor on a supply more than 1% unbalanced should be derated, run below its rating, to stay within its temperature limits. At 3%, the derating is about 10%.
3. Higher maximum demand. The extra losses add to the kW drawn in every demand interval, so they raise the demand reading too.

So the phase voltages are measured, the imbalance is calculated by the meter or a power quality analyser, and the acceptable limits and derating factors come from the standards, not from the meter.

06Temperature: the off-meter variable

What the meter does not measure

Unlike every parameter above, temperature is not measured by the electricity meter. Yet it is one of the most significant factors in how much energy a facility actually uses, and therefore in what the meter registers.

How temperature affects your kWh

The pointTemperature is entirely off-meter. The utility does not measure it, and no standard requires it to, yet it is a root cause of higher losses in conductors, transformers and motors, and the meter captures those losses as kWh.

That is why the Arrhenius rule matters: every 10 °C reduction in operating temperature halves the rate of insulation ageing, and cuts resistive losses too. Our article on the Arrhenius rule sets out the full analysis.

07Other key inputs and billing variables

Frequency

Measured directly, from the zero crossings of the voltage waveform. Grid frequency is tightly regulated, at 50 Hz in the UK and Europe and 60 Hz in North America, and rarely moves enough to affect billing, but the meter uses it to synchronise its sampling and its FFT.

Maximum Import Capacity

A contractual parameter, not a measurement: the maximum demand, in kVA or kW, that the distribution network operator (DNO) has agreed to supply. If measured maximum demand exceeds it, an excess capacity charge is levied. It is purely a tariff mechanism.

Time-of-use periods

The meter’s clock assigns each kWh and kW reading to a time-of-use band, such as peak, off-peak or shoulder. The bands are programmed into the meter from the tariff schedule: contractual, not measured, though they set the rate applied to each unit.

DUoS time bands

In the UK, Distribution Use of System charges vary by time band: red, amber and green. The meter records consumption in each band, and the network operator’s published tariff sets the charges. They are derived from the tariff, not from a measurement.

08Measured or standard-derived: the complete picture

Exhibit 7 brings every input together: what the meter measures, what it calculates and under which standard, and what comes from the tariff or the contract.

09How long power quality improvements take to reach the bill

This is the critical question. The answer depends on which part of the bill you are looking at, because each has a fundamentally different time lag.

kWh: fewer from the moment power quality improves

Immediate

The first demand interval to record a lower average

15–30 min

The first full billing period: every charge but a ratcheted demand

1 month

The longest a demand ratchet can hold billed demand up

12 months

kWh, energy consumption

Immediate. kWh is a continuous integral, so the moment power quality improves, whether through better power factor, lower harmonic distortion or better-balanced phases, the meter registers fewer kWh from that instant on. There is no lag. The next bill reflects the lower consumption for whatever part of the month the improvement was running.

kW demand, within the current billing period

One demand interval, 15 to 30 minutes. After the improvement is switched on, the very next interval records a lower average. But the maximum demand for that billing period is already set by the highest interval before it, so the full demand benefit arrives in the first full billing period afterwards, usually the following calendar month.

Worked exampleIf a site’s peak demand of 800 kW fell on day 3, and a correction system is installed on day 15, its billed demand for that month stays at 800 kW.

kW demand, under a ratchet clause

This is where the lag becomes significant. Many commercial and industrial tariffs include a demand ratchet: a clause that sets the minimum billable demand at a percentage of the highest peak in the previous 11 or 12 months.

Worked exampleOn an 80% ratchet, a site whose demand once peaked at 800 kW is billed for at least 640 kW in each of the following 11 months, whatever it actually draws, until that peak drops out of the window.

Exhibit 8 sets out the lag for each part of the bill.

The kWh benefit is immediate. The kW benefit can take up to twelve months to arrive in full, and many facility managers don’t realise their demand ratchet is the reason.

The complete timeline

For a facility whose power factor is corrected to 0.98+, with its harmonic distortion and phase imbalance reduced, the benefits arrive in this order:

1. At once kWh savings begin immediately. The next demand interval, 15 or 30 minutes, records a lower average, and the power factor the meter calculates rises to 0.98+.
2. The first bill The kWh saving shows for the part of the month since the improvement, and power factor and reactive charges fall. Maximum demand may still reflect the old peak, if it came earlier in the period.
3. The first full billing period The first period that runs entirely at the improved power quality. kWh, kW demand, power factor and reactive charges all reflect it in full, and power factor penalties fall away, unless a demand ratchet applies.
4. Month 12, under a ratchet The ratchet lifts once the old peak has dropped out of the trailing 11-month window. From then on, billed demand follows the improved demand, with no floor.