SOLAR PV

Solar PV Commissioning Checklist: IEC 62446-1 Tests

Commissioning proves a PV system matches its design and is safe to energise. Here are the tests in order, a worked string sum and a filled test record, using invented figures.

By EDWartens engineering team 19 October 2026 13 min
Solar PV Commissioning Checklist: IEC 62446-1 Tests

Commissioning a grid-connected solar PV system means proving the installation matches the design and is safe before it is energised. IEC 62446-1 sets out the documents, inspection and tests: polarity, string voltage and current, insulation resistance and earthing. This guide gives the order, a cold-day string sum and a filled test record.

One fact about our own offer, so you can judge this fairly. The EDWartens solar and EV charging engineer job simulation is a training project at a simulated company, done day by day with project documents, emails and tickets. It is not employment, an internship or a work placement. The first week is free and every ticket is checked as you hand it in. A graded, verifiable certificate is optional, and the longer Premium version is paid; prices show in your own currency before you pay. It shows on your EDWartens profile under Work experience, marked as a job simulation.

Safety first. Testing a live PV array involves direct current that cannot be switched off while the sun shines. Only competent people, with the right instruments, permits and procedures, should do it. This guide teaches the logic of the records and is not instruction for working on a live system. Read the HSE guidance on electricity at work if you work in Great Britain.

What does IEC 62446-1 cover?

The standard covers grid-connected PV systems. It says what information must be handed over, what an installer inspects, which tests to carry out and how to record them. It does not replace national wiring rules or the network operator's requirements, and a system with batteries needs more than this document alone. The IEC page for the standard lists the published edition. At the time of writing, in October 2026, the edition on sale is the 2016 one, with an amendment from 2018, and a revision is in preparation. Editions change, so check which one your client's specification names before you build a record around it.

The standard is protected by copyright and is not free to read. This guide describes the steps in our own words and does not reproduce its text or tables. Use a licensed copy for the exact limits.

Documents to hand over

A system is not finished when it works. The client needs the paper that proves what was built. A typical handover pack holds:

DocumentWhat it showsCommon gap
System detailsRated power, module and inverter types, installer, dateMissing serial numbers
String layout and array drawingWhich modules are in which string and where it connectsLayout does not match the roof
DatasheetsModule and inverter data used in the designOld datasheet revision
Single-line diagramProtection, isolators, metering, earthingNot marked as built
Test resultsEvery measurement, with date, instrument and conditionsResults with no conditions recorded
Operation and maintenance informationShutdown, isolation and cleaning guidanceLeft out of the pack

Before you test: check the design sum

Testing proves the installation matches the design, so first check the design is sound. The classic sum is how many modules go in a string. Here is a worked example with invented module and inverter data.

QuantityValueNote
Module open-circuit voltage at standard test conditions (Voc)37.2 VFrom the datasheet
Voc temperature coefficientminus 0.29 percent per kelvinFrom the datasheet
Coldest expected cell temperatureminus 5 degrees CSite design assumption
Inverter maximum DC voltage1,100 VFrom the datasheet
Design limit95 percent, or 1,045 VMargin chosen by the designer

The temperature rise from the standard 25 degrees C to minus 5 degrees C is 30 K. The cold-day Voc of one module is:

37.2 V x (1 + 0.0029 x 30) = 37.2 V x 1.087 = 40.44 V.

Divide the limit by that figure: 1,045 V / 40.44 V = 25.8 modules. Round down, which gives a maximum of 25. A designer who picks 24 modules leaves a small margin for tolerance. Check the other end too, because 26 modules would give 1,051 V on a cold morning. That is below the 1,100 V rating but above the chosen design limit, so it fails. Then check the other direction: in summer heat the string voltage falls, and it must stay above the inverter's start-up and tracking minimum. A string that is fine in winter and drops out in July is a common finding.

The test sequence

The order matters because each test assumes the one before it passed. The sequence below paraphrases the structure of the standard and the usual practice. Your copy of the standard has the exact requirements.

OrderTestWhat it provesWhat a problem looks like
1Visual inspectionComponents, labels, mounting and cable routes match the designDamaged cable, missing label, wrong module type
2Earth and equipotential bonding continuityProtective earthing and bonding are connected throughoutHigh resistance joint, missing bond between frames
3PolarityEvery string is connected the right way roundReversed string, which can damage equipment
4String open-circuit voltageRight number of modules, no open connectionsReading about one module short or irregular between strings
5String short-circuit currentEach string produces healthy current for the lightOne string clearly lower than its neighbours
6Insulation resistance of DC circuitsCables and modules are insulated from earthReading below the limit, often in damp conditions
7Functional testsIsolators, inverter and protection work as designedInverter does not start, trip settings wrong
8Performance checkOutput is plausible for the light and temperatureOutput far below expected

The standard also describes extra tests, such as a full current-voltage curve and infrared imaging, that some clients ask for. They add information on module health but are not a substitute for the basic set above.

A filled string test record

Here is an invented record for an array of eight strings, each of 24 of the modules above. Conditions on the day: back-of-module temperature 22 degrees C, irradiance 640 W/m2 from a reference cell, steady for the whole set.

First work out what to expect. Open-circuit voltage: one module at 22 degrees C is 37.2 V x (1 + 0.0029 x 3), which gives 37.52 V, and 24 modules give about 900.6 V. Short-circuit current scales with light: 10.9 A x 640 / 1,000 = 6.98 A. The record uses a plus or minus 2 percent window on voltage (882.6 to 918.6 V) and plus or minus 5 percent on current (6.63 to 7.33 A), invented limits for this example. Your client's specification sets the real ones.

StringModulesVoc measured (V)Isc measured (A)Insulation at 1,000 V (megohms)Decision
S1249017.0210Pass
S2248996.9185Pass
S3249027.0240Pass
S4248636.9160Hold: Voc low
S5249007.10.38 (rain)Hold: insulation low
S6249036.9205Pass
S7248987.0198Pass
S8249005.2175Hold: Isc low

Three strings need a decision, and each tells a different story.

S4, voltage low. The reading is 37.6 V under the 900.6 V expectation. That is almost exactly one module. Before any theory, count the modules in the string and compare with the layout drawing. The likely finding is a missing or bypassed module or a loose connection, and the fix is physical, not a recalculation. A single module with a failed bypass diode would drop the voltage by about a third of a module, which is not what the reading shows.

S5, insulation low after rain began. The reading of 0.38 megohms is far below what the other strings show and under the limit for this system. Record the conditions. Do not sign it off because it is raining, and do not condemn the string on the first reading either. Re-test when dry. In this invented case, the dry retest the next morning gives 24 megohms after a wet connector was dried and re-seated. Both readings stay in the record with the cause and the action, which is far stronger evidence than a single clean number.

S8, current low. 5.2 A is about 25 percent under the 6.98 A expectation, while the voltage is right. That pattern points to shading, soiling, a cracked or mismatched module, or a poor connection, not a missing module. Check the reference cell, look for shadows from roof equipment, inspect the connectors and consider an infrared check. Retest when conditions are steady and write the cause in the record.

The skill the record tests is not arithmetic. It is the habit of comparing every reading with an expectation you worked out first, and refusing to mark a result as passed because it is probably fine.

Safe energising: a checklist

  • All test results recorded with date, instrument, conditions and the name of the tester.
  • Every hold from the record closed out, with the retest result next to it.
  • DC and AC isolators labelled and working, with warning labels in place.
  • Earthing and bonding confirmed against the single-line diagram.
  • Polarity confirmed on every string before any inverter connection.
  • Inverter settings, protection and export limits checked against the approved connection.
  • Cable routes, glands and fire barriers inspected and sealed.
  • Emergency shutdown information visible at the main supply and the inverter.
  • Roof access and fall protection controlled, and no one working under test.
  • Network operator notification or approval confirmed as required.
  • Lockout and tag procedure used for any work after the first test.
  • Handover pack complete and a copy given to the client.

Grid connection: G98 and G99, and other regions

In Great Britain, generators connect under the Energy Networks Association's procedures. Generation of up to 16 A per phase usually follows G98, with notification after installation, and larger generation follows G99, which needs approval from the network operator first. Documents are updated, so check the Energy Networks Association for the current version. India and Gulf states have their own utility rules and standards, set by regulators and utilities, with policy from bodies such as the Ministry of New and Renewable Energy. The IEA PVPS programme publishes international reports on PV systems. Never assume a rule from one country applies in another.

Adding EV chargers on a limited supply

Chargers often join the same site after the PV is built, and the supply is the limit. Take an invented example: a site with a 100 A per phase supply and a base load that peaks at 38 A per phase, wanting four 22 kW chargers (32 A per phase each). Together they could draw 128 A, far above the 62 A of headroom (100 A less 38 A). Without management the main fuse would trip or the supply would be overloaded.

A load management controller shares the headroom. With the base load at its peak, the four chargers share 62 A, which is 15.5 A each, about 10.7 kW on a 400 V three-phase supply. When the base load drops to 20 A, the headroom is 80 A and each gets 20 A. The rule is simple: the controller measures the site load, subtracts it from the supply limit, divides the rest among the active chargers and never goes below the minimum current an electric vehicle accepts, which is 6 A. Four chargers at 15.5 A is comfortably above that. Our post on the solar and EV charger commissioning simulator shows how this looks in a lab.

Practise it

To practise the sequence, sums and records with feedback, try the solar and EV charging engineer job simulation. It is a project at a simulated company, with design checks, a test record and a connection application. The electrical design engineer simulation, the substation and motor control centre engineer simulation and the maintenance electrical engineer simulation cover the neighbouring skills, and all sit on the job simulation hub.

For hands-on practice, the solar and EV charger simulator and the substation and motor control centre simulator are in the EDWartens lab, and the multimeter simulator lets you practise reading instruments. For isolation practice, read our guide to safe isolation procedure, and for the wider field, our post on renewable energy automation with PLC and SCADA. Our free courses on electrical fundamentals for technicians and industrial motor control build the basics.

A finished simulation shows on your EDWartens profile under Work experience, marked as a job simulation, and the optional certificate has a public page where anyone can verify it. The electrical design internship is a free guided four-week option, and the jobs board lists current openings.

Last updated 19 October 2026. All figures in the worked examples are invented. Edition status, limits and utility rules change, so work from the current documents.

Practise with us

Internship

Electrical Design

Free internship: 4 weeks, a free certificate and nothing to buy. Premium adds mentor calls and the graded certificate.

See the internship

Lab

Panel Lab

Practise in your browser. Free, nothing to install.

Open the lab (opens in a new tab)

Roles

Electrical design

Roles that ask for this skill, 72 open in this country now. Free to see and to apply. We are not a recruitment agency.

See roles

Take the free course

Questions

What is IEC 62446-1?

It is the international standard for the documentation, commissioning tests and inspection of grid-connected photovoltaic systems. It says what records must exist, what to inspect and which tests to run and record before a system is handed over. National rules and your client's specification apply on top of it.

Which tests are done first on a PV array?

Visual inspection and polarity come first, because every later measurement assumes the wiring is correct. Then open-circuit voltage and short-circuit current on each string, then insulation resistance, with earth continuity checked before the array is touched. Follow the order in the current edition of the standard.

Why measure string open-circuit voltage and short-circuit current?

Open-circuit voltage confirms a string has the right number of modules, connected the right way. Short-circuit current shows whether each string produces a healthy current for the light on the day. Both are compared with the expected value, corrected for temperature and irradiance, and a spread between strings is the clue.

Can I do an insulation resistance test in the rain?

Wet connectors and damp cable can lower the reading, so record the conditions and retest when it is dry. Do not sign off a failure just because the weather could explain it, and do not ignore a low reading either. A retest that passes dry, with the cause found and fixed, is the evidence you need.

How do I calculate the maximum number of modules in a string?

Correct the module open-circuit voltage up for the coldest expected cell temperature using the voltage temperature coefficient, then divide the inverter maximum DC voltage, less your design margin, by that figure. Round down. Also check the minimum number needed to stay above the inverter's start-up voltage in hot weather.

What is the difference between G98 and G99?

In Great Britain they are the Energy Networks Association procedures for connecting generation to the network. Smaller generation, up to 16 A per phase, normally follows G98 and a notification after installation. Larger generation follows G99 and needs the network operator's approval first. Check the current documents.

Is the IEC 62446-1 standard free to read?

No. The IEC and national standards bodies sell it, and the text is protected, so this guide paraphrases and points you to the source. Your employer or college may hold a licensed copy. Always work from the current edition and the version your client's specification names.

Does a job simulation replace site experience?

No. It teaches the sequence, the sums and the records, and checks your answers. Real testing must be done by competent, authorised people with the right instruments and permits. Describe a simulation honestly as training at a simulated company.

Sources

Written by the EDWartens engineering team for general education. Product names are trademarks of their owners; mentioning them does not imply endorsement. Prices and terms of other providers were checked on the date shown and can change.