Electric Motor Running Cost Calculator
A motor can weigh heavily on a plant's electricity bill when it runs several thousand hours a year. This calculator estimates its annual running cost, and the potential saving from a variable speed drive (VFD) that matches the delivered power to actual demand, instead of running the motor flat out with a valve or damper throttling the flow.
On a centrifugal load (pump, fan), the saving follows the affinity laws: power varies with the cube of speed — cutting speed by 20% doesn't cut power by 20%, it cuts it by roughly 49%. On a constant-torque load (conveyor, screw compressor, hoist), power simply varies with speed: the saving is proportionally much smaller.
The motor and how it's used
Result
This motor, without a drive, costs roughly 7,258 €/year to run at this load point.
- Annual cost without a drive 7,258 €
- Annual consumption without a drive 48,387 kWh
| Annual consumption — without a drive | 48,387 kWh |
| Annual cost — without a drive | 7,258 € |
The calculation, in detail
Annual consumption = Power × Load factor × Hours ÷ Efficiency
Centrifugal load: P(drive) = P × (Speed %)³
Constant-torque load: P(drive) = P × (Speed %)
Deriving the affinity laws: why power varies with the cube of speed on a pump
A centrifugal pump or fan is a spinning machine shoving a fluid. Flow Q scales with impeller speed N (same machine, same system, no static head to argue about). Head H — the pressure you have to push against — scales with N², because dynamic pressure is ½ρv² and v follows the tip speed. Hydraulic power is Q × H × density / efficiency. Put N and N² together and power goes as N³.
Q ∝ N · H ∝ N² · P ∝ N³
80 % speed is not 80 % power. It is 0.8³ = 51 % of the shaft power, before you even talk about the motor. That is why a VFD on a half-closed damper looks like witchcraft on the bill, and why this form has a cube on the centrifugal setting. The cube is not a marketing trick. It is Q × H.
Static head breaks the story. A pump lifting into a tall tank still has to make that height at any flow. The system curve is not a pure N² parabola, so P is no longer a pure N³. If most of the head is static, the saving from slowing down is much thinner than the cube promises. Type the cube anyway and you will buy a drive that never earns its keep.
Valve/damper throttling vs a drive: where the lost energy actually goes with throttling
Throttling is running the machine at full speed and then throwing away the extra flow as turbulence in a valve or a damper. The motor still sees nearly rated torque (the fan is still spinning at Nrated), so it still draws nearly rated kW. The missing flow did not vanish from the universe — it became noise and heat in the restriction. You paid to create pressure, then paid again to destroy it.
A VFD lowers N so the machine only makes the flow you want. On a centrifugal load the shaft power falls with the cube; the motor input follows, minus the drive's own few percent. The energy you used to dump in the valve is simply not drawn from the grid. That is the whole difference. On a constant-torque load (conveyor, positive-displacement pump, hoist) there is no cube: power tracks speed, so a 20 % speed cut is roughly a 20 % power cut, and throttling was never the alternative anyway.
Payback period of a variable speed drive
This page gives you an annual saving. Payback is the drive installed cost divided by that saving. Same arithmetic as the motor-replacement calculator, with the VFD quote in the “purchase” slot and no grant unless you have one.
Payback (months) = installed drive cost ÷ annual saving × 12
A 15 kW centrifugal duty, 4,000 h, 80 % speed: the cube has already halved the shaft power, so the annual euros are not small. A constant-torque duty at the same 80 % is a much slower payback — sometimes slow enough that you fit the drive for process control and treat the kWh as a side-effect. Do not forget the extras in the quote: filtered cables, an EMC cubicle, a dV/dt filter on a long motor lead. They belong in the numerator.
Harmonics and power quality: the technical trade-off to know before installing a drive
A VFD is a rectifier plus an inverter. The rectifier draws current in pulses, not a sine. Those pulses are harmonics: 5th, 7th, 11th… They heat transformers, confuse power-factor relays, and can trip a residual-current device that was perfectly happy with a DOL motor. A single small drive on a stiff supply is usually fine. A hall full of unfiltered six-pulse drives is not.
- DC choke or AC line reactor — cheap, knocks the worst of the current distortion down.
- 12-pulse or active front end — for large ratings, or a weak supply.
- dV/dt or sine filter — protects the motor winding on a long cable, not the supply.
The drive also adds 2–3 % of loss at full load, more at very light load. This form lets you type that efficiency. Leaving it at 97 % is honest for a decent unit at reasonable load; 99 % is wishful.
Near-constant-speed applications: when a drive brings almost nothing
If the machine has to run at 98 % of rated speed all day, the cube of 0.98 is 0.94. A 6 % shaft-power cut, minus the drive loss, is a rounding error. Mixers that must not slow down, hydraulic power packs on pressure, saws, a compressor already on load/unload with a small receiver — look at the speed field. If you cannot type a number comfortably below 90 %, the VFD is a starter and a soft-ramp, not an energy project.
There is still a case for the drive: starting current, mechanical shock, a future process change. Just do not sell it on kWh you will not see.
Real load duty cycle vs average load factor: the limits of this approximation
The form uses one load factor and, with the drive on, one speed. Real plant is a timetable: 2 h at 50 % speed, 6 h at 80 %, 1 h at 100 %, nights off. The cube is not linear, so the energy of that timetable is not the cube of the average speed. Average 80 % could mean “always 80” (P ∝ 0.51) or “half the time 60, half the time 100” (P ∝ 0.5×0.216 + 0.5×1 = 0.61). Same average, 20 % more energy in the second case.
Anyway, if you have a logger, bin the hours by speed (or by flow) and sum P(N) × hours. If you do not, this page is an order of magnitude — good enough to decide whether a quote is interesting, not good enough to put on a board paper as a guaranteed saving.
Constant-torque loads: no cube, still a VFD sometimes
Conveyors, extruders, hoists, screw compressors: torque stays roughly constant as speed drops, so power tracks N, not N³. The form switches the exponent from 3 to 1 when you pick that load. The saving is real if you can actually spend hours at reduced speed. If the conveyor only ever runs at line rate, you are back in the “near-constant-speed” case above.
A VFD still earns its place as a starter: current limited to around 1.5–2 × In instead of 6–8, torque you can ramp, and braking if you pay for it. That is a star-delta conversation, not an affinity-law conversation.
Minimum speed, motor cooling, and a shaft that still has to dump heat
The fan on a TEFC motor is on the shaft. Slow the shaft and you starve the cooling. Many manufacturers want a minimum speed of 20–30 Hz on a self-cooled motor at rated torque, or a separate blower. A centrifugal pump at 40 % speed is usually fine (torque has fallen with N²). A constant-torque conveyor at 20 Hz may cook the motor with a perfectly legal VFD setting.
The cube assumes a variable-flow centrifugal machine on a system curve with little static head, and a motor still cooled at the speed you typed. A boiler feed pump, a submerged mixer, or a 10 Hz crawl at rated torque will not match this page. Check the system curve and the motor's speed/torque thermal limit before you treat the annual saving as cash.