Variable Frequency Drives (VFD) in Pumping: When It Pays

Executive summary

The variable frequency drive (VFD) is the most-cited efficiency lever in pumping, but it is not always the right call. Its value comes from the affinity law: power falls with the cube of speed (P ∝ n³), so trimming the revolutions a little saves a lot. This article explains when a VFD pays for itself and when it does not, what sets it apart from throttling with a valve, which selection cautions avoid a costly failure (harmonics, motor derating, minimum speed), and closes with a payback example worked step by step.

Who this is for

For plant, energy and maintenance engineers deciding whether to fit a drive to an existing pump, and for anyone specifying new equipment who must choose between speed control, throttle control or start-stop control. If you have to justify the investment to management — or rule out a VFD that will not deliver — here are the technical criteria to do it with numbers rather than intuition.

The real plant problem

Most centrifugal pumps are selected for the worst case and then run almost always below it. To lower the flow, the most common practice is to close a valve: the pump keeps spinning at full speed and the surplus energy is dissipated as loss across the valve. It is the hydraulic equivalent of driving with the accelerator floored and controlling speed with the brake. The motor draws almost the same power whatever flow it delivers, and that difference — invisible on the panel — is paid for every hour of the year. The VFD attacks exactly there: instead of throttling, it matches speed to demand.

Engineering fundamentals

The affinity laws relate flow (Q), head (H) and power (P) to shaft speed (n):

Q ∝ n  ·  H ∝ n²  ·  P ∝ n³

The practical consequence lives in the third relation: because power depends on the cube of speed, a 15% cut in revolutions drops power to 0.85³ = 0.614, a saving near 39%. Throttling with a valve takes no advantage of this: it keeps the pump at full speed. There is, however, a decisive piece of fine print: the cube law holds cleanly when the system curve is friction-dominated and passes through the origin. If there is significant static head (lifting to an elevated tank), lowering the speed drops the flow far sooner than the formula suggests and the real saving is smaller. That is why diagnosing the system comes before buying the drive.

How to evaluate it step by step

• 1. Operating profile: measure or estimate how many hours a year the pump works at each flow. A VFD only pays if it spends real time below 100%.
• 2. Nature of the system: separate static head from friction head; the more friction dominates, the more the drive delivers.
• 3. Power at each speed: apply P ∝ n³ to estimate the VFD power at each point of the profile.
• 4. Energy saved: subtract the VFD power from the power of throttling at full speed and multiply by the hours.
• 5. Selection cautions: review harmonics, motor derating at low speed and the allowable minimum speed (see below).
• 6. Payback: divide the total investment (drive, installation, filters) by the annual money savings.

Worked example with numbers

A pump whose motor draws 37 kW when throttled at full speed. Average real demand is 85% of the flow, so with a drive it runs at 85% speed. Operation of 5,000 hours/year and a tariff of $0.14/kWh. Total installed drive investment (with line filter): $12,000.

The message is clear: a cut of just 15% in speed removes almost 39% of the power, because the cube law rewards every point. The drive recovers $9,996 a year and pays for itself in a little over a year; after that, it is net savings. If demand were even more variable — more hours below 85% — the benefit would grow further still.

When it applies and when it does not

It pays in variable-flow systems dominated by friction: process pumping, pressurized water distribution, recirculation, cooling towers. There the VFD replaces throttling and harvests the cube law. It delivers little when the head is almost purely static (lifting to an elevated tank): reducing speed lowers the flow before the power and the saving evaporates; it can even push the pump outside its recommended range. Nor is it worth it if the pump runs always at 100% (constant flow): without demand variation there is no energy to recover, and the drive only adds losses and cost. At constant flow, a high-efficiency motor and good selection outperform a VFD.

Common mistakes

• Ignoring static head: applying P ∝ n³ to a static system overestimates the saving and disappoints on the bill.
• Forgetting motor derating: at low speed the motor's own fan cools less; for continuous operation at low frequency you must derate (NEMA MG-1) or use forced ventilation.
• Not anticipating harmonics: drives inject harmonic currents into the grid; on large installations you must evaluate filters or reactors to comply with IEEE 519.
• Going below minimum speed: below ~40–50% of speed many pumps lose useful head, seal cooling or lubrication; set a floor.
• Insulation stress from long cable: PWM pulses over long cables between drive and motor create voltage spikes; use a dV/dt filter or an inverter-duty motor.
• Pricing only the drive: real payback includes installation, filters, engineering and commissioning, not just the equipment price.

Decision checklist

• Does the pump spend real hours below 100% flow?
• Is the system friction-dominated rather than static-head?
• Did you calculate power with P ∝ n³ and not linearly?
• Did you define a safe minimum speed for the pump and the seal?
• Did you evaluate harmonics, motor derating and cable length?
• Does the total investment — not just the drive — pay back from the annual saving?

Frequently asked questions

When is it worth installing a variable frequency drive on a pump?

It is worth it when the pump runs at variable flow and the system is friction-dominated: there, reducing speed cuts power with the cube (P ∝ n³) and the savings are large. It is not worth it if the flow is constant or the head is almost purely static, because the cubic saving barely materializes.

What is the difference between a VFD and throttling with a valve?

Throttling keeps the pump at full speed and dumps the surplus energy as loss across the valve, so consumption barely drops. The drive lowers shaft speed to deliver exactly the flow needed, and by the affinity law power falls with the cube of speed: it is the difference between braking and easing off the accelerator.

What should you watch for when selecting a variable frequency drive for a pump?

Check motor derating at low speed due to cooling, define a safe minimum speed for the pump and the seal, evaluate harmonics fed back to the grid (filters or reactors, IEEE 519), and if the cable between drive and motor is long, protect the insulation with a dV/dt filter or an inverter-duty motor.

SEMHYS tools

To put a number on your case, our free energy and savings calculator estimates the consumption in kWh, the saving by flow range and the payback of a drive. Before investing, it is wise to confirm the pump operates near its best efficiency point with the pump calculator: sometimes the first saving is not the VFD but correcting the selection. And if you are looking for efficiency equipment and accessories, check the store.

References

1. U.S. Department of Energy & Hydraulic Institute (2004). Variable Speed Pumping: A Guide to Successful Applications.
2. U.S. Department of Energy. Adjustable Speed Pumping Applications — pump system tip sheet (affinity laws and variable speed).
3. IEEE 519-2014. Recommended Practice for Harmonic Control in Electric Power Systems.
4. NEMA MG-1. Motors and Generators — application of motors with drives and speed derating.