Building a Defensible Power Quality Baseline (IEC 61000-4-30 and EN 50160)

Two situations call for a power quality baseline. In the first, a facility and its electricity supplier disagree about a recurring disturbance — the site says the supply is poor, the supplier says the site’s own loads are to blame. In the second, a new process, building or major machine is about to be commissioned, and everyone needs to know what “normal” looked like beforehand.

In both cases, informal evidence fails. A screenshot of a dip proves that something happened, but not how representative the monitoring period was, how complete the data is, or which method produced the numbers. This guide sets out how to build a baseline that survives scrutiny.

The question to answer

What quality of supply was delivered, and is the evidence comparable? “Comparable” is the important word: comparable with the contract, with a later survey, and with the other party’s measurements.

Method and limit are two different things

Two standards come up in almost every baseline discussion, and they do different jobs.

Keep them separate in every report: first state how the data was measured, then state which limit it was compared against, and why that limit applies.

Agree the rules before you record

Most baseline disputes are lost before the first sample is taken, because the parties never agreed what they were measuring. Settle these points first:

1. The measurement point — usually the point of common coupling, or the facility incomer.
2. The declared voltage and the wiring topology (single-phase, three-wire delta or four-wire star).
3. The reporting period. A week is the common minimum because it covers weekday and weekend load patterns.
4. Event thresholds and settings — what depth and duration counts as a dip, a swell or an interruption.
5. The acceptance rules that will be applied afterwards, and who agreed them.

Document the instrument, not just the data

A baseline should let another engineer reproduce your conclusion. Record:

• device and firmware identity, and calibration status,
• probes and CT ratios,
• how time was synchronised and how much uncertainty that leaves,
• every data gap, and why it happened.

Inspector V4 is designed for permanent installation at an incomer, with two isolated channels, a measurement range up to 1000 V RMS (CAT III 1000 V), a 24-bit converter sampling at 32 kS/s per input, and a six-hour internal battery so the record continues through an interruption. The Portable V4 carries the same measurement engine for survey work.

Reading the evidence

Separate the two kinds of result:

• Continuous characteristics — voltage magnitude, frequency, unbalance, harmonics and flicker (Pst, Plt) — summarised over the agreed aggregation intervals and compared against the agreed limits.
• Discrete events — dips, swells, interruptions and transients — listed with timestamp, depth and duration, and related to operational complaints.

Include exclusions and invalid intervals in the summary rather than quietly dropping them. A baseline that admits a six-hour data gap is more credible than one that hides it.

The report

A repeatable baseline report contains:

1. the monitoring boundary and the agreed rules,
2. data coverage and every exclusion,
3. trends for each continuous characteristic,
4. the event list,
5. a small number of representative waveforms,
6. the instrument and configuration record.

Preserve the raw evidence and the configuration so the conclusion can be reviewed — by the supplier, a consultant or your own team a year later.

What success looks like

An auditable baseline that supports commissioning, supplier discussions and repeat surveys. Useful indicators:

• valid data coverage over the period,
• the number of events that can be attributed to an operating cause,
• time taken to close a complaint,
• whether a later survey reproduces the result.

Related reading

• What a power quality analyzer actually tells you
• Harmonics and poor power factor
• Surveying a site before adding the next major load