A power quality assessment for a solar farm proves the plant keeps harmonics, flicker, voltage unbalance and voltage changes inside the limits its network service provider (NSP) has set. It runs in three stages: a pre-connection emission study, baseline logging at the connection point before energisation, and compliance measurement once the plant is exporting. The output is a report the NSP reviews and signs off before the power quality conditions in the connection agreement are closed.
The study asks whether the design will comply, the baseline records what the network was already doing, and the compliance test shows what the plant actually does. Skip one and the NSP will find the gap.
This article covers that test sequence. How inverters create harmonics is explained in our solar farm harmonic analysis guide, and how networks set and allocate limits in our guide to AS/NZS 61000.3.6 harmonic emission limits.
Where the Power Quality Obligations Come From
Power quality limits for a solar farm come from two places, and which one dominates depends on the size and registration status of the plant.
Registered generators: the GPS and connection agreement
For a plant registered with AEMO, power quality sits inside the generator performance standards. The National Electricity Rules include a quality of electricity generated standard within Schedule 5.2, which links to the Australian power quality standards: AS/NZS 61000.3.6 for harmonics, AS/NZS 61000.3.7 for flicker and AS/NZS 61000.3.13 for unbalance. The agreed performance standard, and the emission limits the NSP allocates under it, end up in the connection agreement.
The important word is allocates. The NSP shares the available headroom at the connection point between existing and future customers and gives your plant a slice, expressed as harmonic voltage or current limits per order. That slice is what the assessment is measured against.
Embedded plant: the DNSP's own standards
A smaller solar farm connected to a distribution network, below the registration threshold, often deals only with the distributor. The limits then come from the DNSP's connection standards and technical requirements, which also draw on the AS/NZS 61000 series but vary in format, measurement duration and reporting expectations between distributors. Many DNSPs also add audio frequency (ripple control) interference requirements, because a plant that loads the ripple signal can disable hot water and load control across a feeder.
Read the connection offer early. Its power quality clauses say what will be measured, where, for how long and against which limits.
Stage 1: Pre-Connection Emission Study
The pre-connection study predicts whether the plant will comply before anything is built. It is done during connection application and detailed design, and it is where most of the engineering effort sits.
Harmonic emission study
The study combines two sets of data:
- Inverter harmonic emission data. The manufacturer's harmonic current spectrum, ideally from type test reports at several output levels rather than a single datasheet figure at rated power.
- Network harmonic impedance. Usually supplied by the NSP as impedance loci, which are envelopes in the R-X plane for each harmonic order covering the range of network configurations the connection point might see.
These go into a harmonic model of the plant, including inverter transformers, the collector cable network, the main transformer and any reactive plant. Cable capacitance in a large collector system can create resonances that amplify particular orders, so the model has to represent the solar farm's electrical design as it will actually be built, not as a single lumped generator.
The study then predicts harmonic voltage distortion at the connection point for each order and compares it with the allocation. If the margin is thin or negative, mitigation is designed at this stage, while it is still cheap to change.
Flicker and unbalance screening
Flicker from a solar farm is rarely a problem in steady operation. The concern is cloud transients, inverter switching and capacitor or filter switching, all of which create step changes in voltage. Screening uses the short-circuit ratio at the connection point and the size of the expected power swings to estimate Pst and Plt and rapid voltage change.
Unbalance screening checks that the plant does not worsen negative sequence voltage. Three-phase inverters are inherently balanced, so the risk usually lies in the network's existing unbalance rather than the plant's own contribution. That is one reason the baseline matters.
Stage 2: Baseline Logging Before Energisation
Baseline logging measures power quality at the connection point before the plant energises. It is the cheapest insurance in the entire regime, and it is often skipped.
Networks are not clean. Background harmonic distortion from loads elsewhere on the network can already be close to, or above, planning levels at particular orders. If the plant is measured after energisation and the 5th harmonic voltage is high, the NSP's first assumption will be that the solar farm caused it. Without a baseline, the developer has no evidence otherwise.
A good baseline:
- Uses a Class A instrument and the same measurement point and transducers as the compliance test will use. Our substation power quality analysis guide covers instrument, CT and VT setup in detail
- Runs long enough to capture weekly load cycles, typically at least seven days
- Records harmonic voltages by order, flicker, unbalance and steady-state voltage, so every compliance parameter has a before-and-after comparison
A baseline is often what settles a disputed result, by showing that the distortion predated the plant.
Stage 3: Post-Energisation Compliance Measurement
The compliance test measures the plant in operation and compares the results with both the allocation and the baseline. It usually forms part of the commissioning and DNSP approval process or, for registered generators, the post-energisation compliance work that sits alongside GPS testing.
Measurement duration
At least seven days of continuous logging is typical, because limits are assessed statistically, commonly at the 95th percentile (the substation guide explains the statistics and data exclusions). Some NSPs ask for longer, or for repeat measurements at higher output stages, so check the connection agreement rather than assuming.
Output levels to capture
Harmonic emission varies with output. Some inverters emit more at partial load than at rated power, and network resonance can make a mid-range output level the worst case. The measurement period needs to capture:
| Operating condition | Why it matters |
|---|---|
| Night and zero output | Shows background network distortion with the plant connected but not exporting |
| Low output (below about 30%) | Inverter emissions relative to fundamental are often highest here |
| Mid range output | Resonance with collector cable capacitance can peak at intermediate levels |
| Near rated output | The condition most studies assume, and the one NSPs expect to see |
| Cloud transients and ramps | Drives flicker and rapid voltage change results |
| Reactive power set points | Filter and capacitor switching, and inverter reactive operation, change the spectrum |
If the plant is being ramped through hold points, the PQ logging needs to be planned around them. A week logged at a 50% export cap does not show compliance at full output, and the NSP may ask for a repeat.
Inverter firmware and settings
Record the inverter firmware version and control settings for the entire measurement period, and freeze them. Firmware updates can change switching frequency, current control loop tuning and harmonic damping behaviour. A firmware change made mid-test, or between the test and commercial operation, can invalidate the result. If the OEM needs to update firmware after the test, expect to discuss re-measurement with the NSP.
When Results Fail
A result above the allocation is not the end of the project, but it needs a clear process.
Confirm the plant caused it
Compare against the baseline and correlate the exceedance with plant output. If the order in question is high at night with the plant not exporting, or does not change with output, the problem is likely background distortion. Present that evidence to the NSP before spending money on mitigation.
Inverter control changes
Many modern inverters have configurable harmonic damping, active filtering functions or adjustable control parameters. Working with the OEM, these can sometimes reduce emission at a specific order without hardware. Any change has to be checked against the plant's dynamic performance, because control changes can affect GPS compliance elsewhere, and models may need updating.
Filter design
Where control changes are not enough, a passive harmonic filter at the main switchboard or high voltage bus is the usual answer. The filter must be tuned against the network impedance loci, not a single impedance value, and checked for new resonances it may create. It also contributes reactive power, which affects the plant's reactive capability and voltage control. Filter procurement lead times can run to months, which is why finding this problem in Stage 1 is far better than Stage 3.
Re-measurement
After any fix, the compliance measurement is repeated in full, with the same instrument, location and coverage of operating conditions. Partial re-measurement is rarely accepted.
What the Compliance Report Contains
The report is what the NSP actually reviews. A complete one leaves nothing to ask for:
Power Quality Compliance Report Contents
- Plant description, single-line diagram and measurement point
- Applicable limits and allocations, with the connection agreement or DNSP standard they come from
- Instrument make, model, class and calibration certificates
- Voltage and current transducer details and ratios
- Measurement period, logging intervals and aggregation method
- Inverter firmware versions and control settings during the test
- Plant output profile across the measurement period
- Baseline results and comparison with post-energisation results
- Harmonic voltages and currents by order against limits, at the 95th percentile
- Flicker (Pst, Plt), unbalance and rapid voltage change results against limits
- Ripple control or audio frequency interference results where required
- Assessment of exceedances, with cause and mitigation
- Conclusion on compliance and engineer sign-off
Plan Your Power Quality Compliance
Power quality problems cost the most when they are found last. A proper study, a baseline and a well-planned compliance test turn a potential hold on export into a routine sign-off.
ElectraGlobe's power quality analysis team runs harmonic emission studies, filter design, baseline and compliance measurement, and NSP reporting for solar farms and BESS across Australia. If you have a connection offer with power quality conditions, or a result you do not trust, talk to our team.
Frequently Asked Questions
Is a power quality test required to energise a solar farm?
Energisation is usually allowed before the compliance test, because the plant has to be exporting to be measured. However, the pre-connection study is normally a condition of the connection offer, and the compliance report is commonly required before the NSP closes out its conditions or releases the plant to full output. Check the power quality clauses in your connection agreement for the exact sequence.
How long is post-connection power quality monitoring?
A minimum of seven days of continuous measurement is typical, because harmonics and flicker are assessed against statistical values over a week. Some NSPs require longer periods or repeat measurements at higher output stages. The connection agreement or DNSP standard sets the actual requirement.
What happens if a solar farm exceeds its harmonic limits?
First, the cause is confirmed by comparing with baseline data and correlating with plant output. If the plant is responsible, mitigation follows: inverter control changes where possible, a harmonic filter where not. The plant is then re-measured. The NSP may restrict output until compliance is shown.
Does inverter firmware affect power quality results?
Yes. Firmware controls switching behaviour and current control, both of which shape the harmonic spectrum. Record the firmware version during testing and do not change it mid-test. A later firmware update can change emissions, so discuss any post-test updates with the NSP before applying them.