Variable speed drives on pumps and fans: when do they pay off?
Published 5 min readLeer en español
Short answer
A variable speed drive pays off on pumps and fans when demand varies, the equipment runs for many hours and flow is currently controlled with a valve or damper. With constant full load, high static head or short running hours, savings are usually small. Measure flow, pressure and power for several weeks before buying.
Key points
- Throttling with a valve or damper wastes energy: the motor keeps pushing while the valve brakes.
- Under the (idealised) affinity laws, power varies with the cube of speed: at 80% speed, roughly 51% of the power.
- Static head limits how far you can reduce speed, and therefore the real saving.
- Check motor insulation, harmonics, cable length, bearing currents and cooling at low speed.
- Work out payback from your real flow and power profile, not catalogue figures.
What does a variable speed drive actually do?
A variable speed drive (VSD, also called a variable frequency drive or VFD) supplies an induction motor with adjustable voltage and frequency. Because the motor’s speed depends on frequency, the drive lets a pump or fan run at exactly the speed the process needs at any moment.
Without a drive, the motor always runs at rated speed. To deliver less flow you have to throttle it: partly close a valve or damper, or recirculate some of the fluid through a bypass.
Why does throttling with a valve or damper waste energy?
When you close a valve, the pump keeps running at full speed and the valve adds an artificial pressure loss. It is like driving with your foot flat on the accelerator and controlling speed with the brake.
The energy the valve “absorbs” turns into heat, noise and wear. With a drive, instead of braking the fluid you reduce the pump speed, and most of that loss disappears.
Fans behave the same way with dampers or inlet guide vanes: they regulate flow, but the motor still works harder than it needs to.
What do the affinity laws say?
The affinity laws describe, in idealised form, how a centrifugal pump or fan changes with speed:
- Flow is proportional to speed.
- Head (pressure) is proportional to speed squared.
- Power is proportional to speed cubed.
That cubic relationship is what makes drives attractive. At 80% speed, ideal power is 0.8 × 0.8 × 0.8 = 0.512, roughly half.
An illustrative example
Suppose a pump draws 11 kW at rated speed and, most of the time, only needs 80% of its flow. It is currently throttled with a valve and, as an assumption, draws about 10 kW in that condition (you would measure this value, not guess it).
- Ideal power with a drive at 80% speed: 11 × 0.512 ≈ 5.6 kW.
- Difference: 10 − 5.6 = 4.4 kW.
- If it runs 4,000 hours a year at that point: 4.4 × 4,000 = 17,600 kWh.
- At an assumed price of €0.15/kWh: 17,600 × 0.15 = €2,640 a year.
These figures are made up purely to show the arithmetic, not a promise. In practice the real saving is lower than the ideal for three reasons: static head, the efficiency of the drive itself (it has losses), and the efficiency of the motor and pump, which falls off at low speed.
Why does static head limit the saving?
The affinity laws hold well in friction-dominated systems, such as closed heating and cooling circuits. But if the pump has to lift water 30 metres to a tank, those 30 metres have to be overcome at any speed.
As speed drops, the head the pump produces falls with the square of speed. At some point it no longer reaches the static head and flow collapses to zero. So in systems with high static head the control range is narrow, and the saving is much smaller than the cube law suggests.
When does a drive pay off, and when doesn’t it?
| Pays off | Doesn’t pay off (or barely) |
|---|---|
| Demand varies through the day or year | Constant load at full power |
| Many running hours per year | Few running hours |
| Currently controlled by valve, damper or bypass | Already runs at its best point without throttling |
| Friction-dominated system (closed circuits, ventilation) | High static head relative to total head |
| Oversized pump or fan | Correctly sized and always at 100% |
| Precise pressure or flow control is needed | Process works fine with simple start/stop |
If your case falls almost entirely in the right-hand column, a drive may still make sense, but not for the energy.
What other benefits does a drive bring?
Even if the energy case is weak, some benefits can justify the investment on their own:
- Soft start and stop. Cuts inrush current and mechanical stress on couplings, belts and pipework.
- Less water hammer. Stop ramps on pumps limit pressure transients.
- Constant pressure control. With a pressure transmitter and a PID loop, the pump holds its setpoint as demand changes.
- Diagnostics. Most drives report current, power, running hours and faults over Modbus or other fieldbuses, which is useful for monitoring.
What should you check before installing one?
A drive is not a plug-and-play component. This checklist avoids surprises:
- Motor insulation. Drive output has fast voltage edges. Older motors may not be designed for it; look for inverter-duty motors or add filters.
- Cable length. Long cables between drive and motor cause reflected-wave overvoltages. Stay within the manufacturer’s limit or fit a dv/dt or sine-wave filter.
- Harmonics. Drives inject harmonics into the supply. With many drives or a weak supply you may need a line reactor or filter.
- Bearing currents. These can damage bearings over time. Mitigate with good earthing, screened cable, shaft grounding brushes or insulated bearings.
- Cooling at low speed. The motor’s own fan turns more slowly and cools less. At low speed and high torque you may need forced ventilation.
- Minimum speed for pumps. Set a minimum so the pump doesn’t run against static head with no flow, overheat, or lose lubrication of seals and bearings.
- Resonances. Some speeds can excite vibration; drives let you skip those frequencies.
- Electrical installation. Power wiring and the mains connection must be done and certified by a licensed electrician.
How do you estimate payback with your own data?
The usual mistake is to estimate savings from nameplate power and a generic percentage. Do it this way instead:
- Measure for several weeks: electrical power, flow and pressure (or at least power and valve position). Include weekdays, weekends and different seasons if demand is seasonal.
- Build the load profile: how many hours the equipment spends in each flow band.
- Establish the system curve: static head plus friction losses. Without it you don’t know how far speed can drop.
- Calculate power with a drive for each band using the pump curves, applying realistic motor and drive efficiencies.
- Multiply by hours and energy price and compare with the total cost: drive, filters, electrical installation, commissioning and any motor replacement.
- Verify afterwards. Keep measuring after installation to confirm the real saving.
If standby consumption across your site is also a concern, our article on idle energy consumption in factories explains how to spot it. And if the drive is part of a wider upgrade, see machine retrofit: modernise or replace.
How we approach it at Kinetifab
We start by measuring power, flow and pressure before proposing anything, and put the payback calculation in writing using your data. We handle control, measurement and integration over open protocols; power wiring is done and certified by a licensed electrician. You’ll find more on our industrial automation page and under energy, and you can reach us via contact.