The duty decides
Membranes are static head
Check the motor
First unit, then repeatWhere the energy saving is real
The affinity laws say that power varies with roughly the cube of speed, which is why a small speed reduction on the right machine produces a large power reduction. The condition attached to that, and it is the condition that gets dropped, is that the system curve has to be dominated by friction.
| Duty | Saving from speed control | Why |
|---|---|---|
| Centrifugal fan, damper throttled | Large | Friction dominated. Remove the damper loss entirely. |
| Centrifugal pump, valve throttled, low static head | Large | Same reason. The control valve was burning the energy. |
| Pump against high static head | Small, sometimes none | Most of the head is lift or pressure, not friction. Slowing down reduces flow long before it reduces power meaningfully. |
| Membrane feed, for example reverse osmosis | Usually none, and risk attached | Required pressure is set by the membrane, which behaves as static head. The pump cannot be slowed without losing the process. |
| Positive displacement | Proportional, not cubic | Flow is proportional to speed. Useful for control, not for a cube-law savings claim. |
On a reverse osmosis high pressure pump the membrane sets the operating pressure, so the duty behaves as static head and the cube law does not apply. A savings case built on affinity laws here will not materialise. And a proposal that recovers the saving by opening the concentrate valve is not an efficiency measure, it is a change to the recovery ratio, with consequences for scaling, membrane life and product quality. If you see that in a proposal, stop.
Where a drive does not pay back on energy it can still be entirely justified, on soft start and mechanical stress, on process control quality, on reduced peak demand, or on removing a starter that is itself obsolete. Those are honest reasons. They are just different reasons, and they should be argued on their own terms.
Retrofitting into an existing installation
The motor has to be checked, not assumed
An older motor may not be rated for inverter duty. Insulation stress from fast switching edges, bearing currents on larger frames, and cooling at reduced speed on a shaft-mounted fan are the three that cause failures six months after a successful commissioning. Shaft grounding, output filtering and forced ventilation are cheap at design and expensive as a retrofit to a retrofit.
Cable length and filtering
Long motor cables produce reflected wave stress at the motor terminals and raise the earth leakage the installation has to tolerate. Decide the filter at design, from the actual measured cable run.
Harmonics belong in the specification
Drives are non-linear loads. On a plant that already has a harmonic problem, a fleet of new six-pulse drives makes it worse, and the symptom shows up somewhere else entirely, as nuisance tripping or transformer heating. Measure before the fleet goes in, and specify the mitigation on the basis of that measurement.
Bypass and failure mode
Decide what happens when the drive fails. A manual bypass starter is cheap insurance on a critical duty and pointless on a non-critical one. What is not acceptable is discovering the answer during a breakdown.
Fleet migrations
Replacing one drive is a job. Replacing thirty across a plant is a programme, and it behaves differently:
- The first unit carries the engineering. Enclosure design, parameter set, control integration and the commissioning procedure are solved once. Subsequent units are a repeat installation at a repeat rate, and the commercial structure should say so plainly.
- Parameter sets are documented and version controlled. Thirty drives configured by hand from memory become thirty different drives within two years.
- Standardise the spare. The point of a fleet migration is that one spare on the shelf covers many machines. That is lost the moment frame sizes are mixed for the sake of a small unit price difference.
- Sequence against production. Which machines can be taken one at a time, which are paired, and which can only be touched at shutdown.
Questions we get asked
Will a VFD reduce our electricity bill?
On a throttled fan or a friction-dominated pump, substantially. On a pump working against high static head, very little. On a membrane feed pump, effectively none. The duty decides it, and it is worth measuring before committing to a fleet.
Can we put a drive on an existing motor?
Often yes, but it has to be checked rather than assumed. Insulation rating, bearing protection on larger frames, and cooling at reduced speed are the three items that decide it.
Do we need a harmonic filter?
It depends on what is already on the supply and how much drive load you are adding. On a small addition to a strong supply, usually not. On a fleet retrofit, measure first and specify from the measurement, because the symptom of getting it wrong appears somewhere other than the drives.
What size enclosure does a drive need?
Larger than the drive. Enclosure height has to be the drive body plus the manufacturer's stated airflow clearance above and below, and that sum should be shown in the design rather than estimated. Undersizing the clearance produces thermal trips that look like drive faults.