Summary Highlights
- Define which instrument functions and parts of the measurement chain are in scope
- Preserve as-found settings, records and measurements before changes
- Separate calibration from adjustment, data checks and event-function tests
- Report uncertainty, decision criteria and untested functions
Article Details
Power quality analyzer maintenance should protect the condition of the instrument, the validity of its measurements and the integrity of its records. These require different checks. A working display or a successful network connection does not establish measurement accuracy; calibration at selected points does not establish every harmonic or event function. Define the scope around the decisions the measurements support.
Define the scope before changing the instrument
Identify the analyzer, firmware, options, voltage inputs, current probes or CT/VT system and any external data collector. Decide whether the work covers the analyzer alone, the installed input chain, the recording functions, the communication path or a combination. Specify the quantities, ranges, waveforms and acceptance criteria to be evaluated.
Preserve the as-found configuration and relevant records before resetting counters, changing ratios, adjusting the clock or updating firmware. Capture the initial symptoms and any flagged or missing data. Changes may affect how earlier records can be interpreted, so document their purpose and timing.
Inspection and safe maintenance
Inspect the case, covers, connectors, display, leads, mounting and safely accessible connections for damage, contamination, moisture or mechanical stress. Check the environmental and supply conditions against the manufacturer’s limits. Connection work, isolation and any retightening must follow the approved site procedure and the manufacturer’s instructions for the actual terminal; there is no universal torque setting.
CT circuits require particular care. Opening the secondary of an energized CT can create a hazardous voltage. Work on measurement circuits, test blocks and earthing arrangements belongs within the authorized site and equipment procedures.
Thermal inspection may identify patterns that deserve investigation at safely visible surfaces. Interpret them with the load, ambient conditions and surface properties recorded. A thermal image alone does not establish a loose connection, an analyzer accuracy error or the condition of inaccessible components.
What is the difference between calibration and verification?
| Activity | What it can establish | Scope to retain |
|---|---|---|
| Calibration | The relationship between reference values and indications, with associated uncertainty | Quantities, points, conditions, method and the equipment included |
| Adjustment or repair | A deliberate change to instrument response or condition | As-found results, the change and the subsequent checks |
| Verification against a requirement | Whether the stated evidence meets a defined requirement | Criterion, decision rule where relevant and measurement uncertainty |
| Connection/configuration check | Agreement of the input model, scaling and channel association with the documented system | Parts checked and inaccessible or unresolved circuits |
| Data-system check | Transfer, timestamps, continuity and interpretation of selected records | Registers, software versions, test period and missing coverage |
A calibration certificate is not automatically a statement of conformity, an adjustment record or a guarantee about future measurements. If adjustment is performed, preserve the initial results and establish the post-adjustment condition. The uncertainty of a reference is part of the evaluation; the observed difference between two displays is not, by itself, the uncertainty of the test.
Plan voltage, current, power and phase tests
Select test points and conditions for the intended quantities and use. Frequency, phase angle, waveform, range and the connection arrangement can affect the result. Use references with suitable calibrated capabilities and uncertainty for the task. Warm-up, environmental limits and permissible errors need the applicable instrument and method specifications.
Compare corresponding quantities over compatible intervals. A handheld reading at a different point or time may be useful for investigation but cannot automatically serve as an accuracy test. Where a conformity decision is required, define the criterion and how uncertainty is considered before interpreting the results.
Include sensors and phase association explicitly
Check probe or CT/VT identity, type, ratio, nominal input and additional software scaling. Confirm that each current channel is associated with the intended voltage channel and reference direction, using the documented wiring model. Ratio and phase effects, operating range, clipping and burden where applicable belong to the chain assessment.
If only the analyzer input was tested, report that boundary. It does not include external transformers, probes, wiring or software that were outside the test. An analyzer’s catalogue accuracy does not describe the complete installed uncertainty.
Investigate negative readings before correcting them
Negative active power can result from real reverse flow, a sign convention or a configuration/connection issue. Check the physical reference direction, phase association, ratios and the manufacturer’s import/export definitions. Reactive power and power-factor signs may use different conventions.
Taking an absolute value, flipping a software sign or reversing a CT without identifying the cause can conceal the original problem. Maintain the as-found evidence and record any authorized correction together with a suitable recheck.
Harmonic and event functions need appropriate tests
A basic voltage/current/power calibration does not establish all dip, swell, interruption, harmonic, transient or synchronization functions. If the work is intended to evaluate functions associated with IEC 62586-2, the specified function tests, applicable edition, equipment, conditions and evaluation method must be identified. A product label or a few steady-state comparison points cannot replace that scope.
Review the supported functions, firmware and options, event thresholds, aggregation, flagging, storage and export behaviour. Check the THD definition and harmonic range for comparisons. A previously stored event or a successful alarm notification does not establish detection of every event or the performance of every input channel.
Validate communications, timestamps and missing data
Keep the physical interface, protocol and register interpretation separate. For RS485/Modbus installations, compare the documented topology and wiring with address, baud rate and parity settings, then verify register address, data type, byte/word order, units, scale and direction. Manufacturer A/B naming, termination and shielding guidance must be checked for the actual system.
Compare local and remote values for the same quantity and time interval. Review communication errors, repeated/stale records, gaps, restarts and buffer limits. A successful response or message acknowledgement does not, by itself, establish a complete end-to-end dataset. Missing values should remain identifiable rather than becoming zero demand or consumption.
Check the time source, clock status, time zone and timestamp definition. Clock writes, firmware updates or configuration downloads can affect records and must follow a planned procedure. Do not assume that every device accepts the same clock register or that enabling synchronization proves its achieved performance.
Set and review the calibration interval
A factory calibration and a suitable operating environment do not, by themselves, establish an exemption from recalibration. Consider the manufacturer’s recommendations, required uncertainty, intended use, use and transport history, environmental exposure, previous calibration results and intermediate checks. Review the interval using documented evidence.
Sections 4–6 of ILAC-G24 / OIML D 10:2022 discuss selecting and reviewing intervals. The guidance does not prescribe one interval for every analyzer. Calibration differs from adjustment, and connection or communication checks do not replace a calibration programme. Any specific applicable requirements need separate consideration.
Report as-found and as-left results without losing context
- Identify the instrument, sensors, configuration, versions and the parts included in the work.
- Retain as-found observations and results separately from adjustments, repairs and as-left results.
- State the method, points, reference equipment, relevant conditions and measurement uncertainty.
- Identify the requirement and decision rule used for any conformity statement.
- Record untested functions, unavailable circuits, missing data and follow-up actions.
As-found describes the condition before intervention; as-left describes the condition at the end of the documented work. A good as-left result does not prove that all previous records were correct. If an initial deviation is significant, assess its possible effect on earlier decisions using the available history. Neither result guarantees future performance or the compliance of the entire facility.
Related Blog Posts
Frequently Asked Questions
Does a calibration certificate cover the complete installed chain?
Only the equipment, functions, ranges and conditions identified in the certificate are covered. External CTs/VTs, probes, wiring and data systems are not included unless the documented scope includes them.
Are calibration and adjustment the same operation?
Calibration establishes a relationship between reference values and indications with associated uncertainty. Adjustment changes the instrument response. Preserve as-found results and establish the post-adjustment condition when an adjustment is made.
Does every power quality analyzer need annual calibration?
There is no single interval for all instruments. Use manufacturer recommendations, required uncertainty, application, environment, transport, previous results and intermediate checks to select and review an interval. Check any specific applicable requirements separately.
Can a basic calibration verify all power quality functions?
A test is limited to the quantities, conditions and functions evaluated. Harmonics, events, timing and other functions require an appropriate defined scope; they are not established automatically by steady-state voltage or current checks.
Should negative active power be corrected by reversing the CT?
First establish the actual flow direction, sign convention, phase association and configuration. Negative power alone does not identify the cause. Rewiring or changing signs without this investigation can conceal the problem.
Does reliable Modbus communication prove the measurements are correct?
Communication checks establish aspects of transfer and interpretation. They do not calibrate the inputs. Scaling, units, register meaning, timestamps, stale values and gaps also need checking for the intended dataset.
Should firmware be updated at the start of maintenance?
Preserve the current configuration, version and relevant records first. Review the reason for the update, compatibility and its possible effects on settings and data, then follow the manufacturer’s procedure. An update is not a universal repair.
What do as-found and as-left results tell the reader?
They separate the initial condition from the condition after the documented work. This helps assess the effect of an intervention and possible earlier deviations. A satisfactory final result does not validate every earlier measurement or guarantee future performance.