In short. Voltage fluctuation covers slow drift, sudden sags and swells, flicker and short transients. Each has a different cause and cure. Classify the event by size and duration from recorded data before choosing a stabiliser, UPS or other fix.
Classify first
| Event | Typical duration | Typical cause | Typical cure |
|---|---|---|---|
| Under or over voltage | Minutes to hours | Weak or lightly loaded supply | Servo stabiliser |
| Sag or swell | Cycles to seconds | Faults, motor starts, capacitor switching | Online UPS or conditioner |
| Flicker | Repeating small variations | Arc furnaces, welders, big cycling loads | Source-side study and mitigation |
| Transient | Microseconds to milliseconds | Lightning, switching | Surge protection |
Measure
Record voltage with timestamps and load. A one-week recording shows daily cycles. See monitoring.
Understand the load effect
- Motors: low voltage raises current and heat; high voltage raises iron loss.
- Drives and electronics: trip on undervoltage or sag.
- Lighting: flicker and reduced life.
- Compressors: stall risk at low voltage.
Source versus site
Ask whether the problem comes from the utility or from your own loads. If your large load causes the dip, correction at the load may not be enough; study the source. See source impedance and reactive power studies.
Chain of thinking
Reading → trend → deviation → diagnosis → intervention → verification.
Illustrative example
A fabrication shop sees evening dips from 415 V to 360 V. The recorder shows the dips when a neighbour’s load rises, a slow event. A servo stabiliser corrects it. Brief sags during welder strikes remain; the controllers go on a small UPS.
Quantifying flicker and rapid changes
Flicker is judged by how much voltage changes and how often. Small repeating changes at certain frequencies are most visible in lighting. Standards define measurement methods for this. A power-quality analyser reports flicker severity values; compare them with the limits in the standard applicable to your supply agreement.
Estimating a voltage dip from a load step
Voltage change (%) ≈ Load step (kVA) ÷ Short-circuit capacity (kVA) × 100
If a 500 kVA machine starts on a supply with 20,000 kVA short-circuit capacity, the dip is about 2.5 percent. If the same machine starts on a weak supply with 5,000 kVA, the dip is about 10 percent. This simple estimate shows why the same plant can behave very differently on different supplies. See source impedance.
Sensitivity of common equipment
| Equipment | Typical vulnerability |
|---|---|
| Variable-speed drives | Undervoltage trip on dips |
| Contactors | Drop-out on deep dips |
| Computers and controllers | Reset on short interruptions |
| Lighting | Visible flicker |
| Motors | Overheating on sustained low voltage |
Mitigation hierarchy
- Reduce the cause: soft starters, staggered starts, drives.
- Strengthen the supply: a dedicated feeder or a higher short-circuit level.
- Protect the sensitive equipment: UPS, ride-through units, voltage-sag support.
- Correct sustained voltage: stabilisers.
Documentation
Keep the event log together with a plant event log. When a trip occurs, check what else was running. A pattern of coincidence is strong evidence of cause.
The master engineering question
How big and how long is the event, and what is the weakest piece of equipment that it hurts?
What to send us
Send recordings and equipment lists. Send your application details.
Questions engineers ask
Is a stabiliser enough for sags?
Not for fast sags. Use UPS or conditioning.
What is flicker?
Repetitive voltage variation that visibly changes lighting.
Can I fix it at the source?
If you own the cause, yes; study it.
How long should I record?
Capture at least a full operating cycle, often a week.
Values are illustrative. Follow applicable standards for limits.