Finding DC and string faults
Finding why a PV string underperforms or trips, from monitoring data to the evidence each test gives. You will learn the order of diagnosis and how to read an I-V curve, and where the safe limits sit.
- Core
- Advanced
- about 16 hours
Written by the GreenHired editorial teamSources last checked 20 November 2026How we research roadmaps
Least intrusive evidence comes first: monitoring data, then a look from a safe position, and tests by authorised people last. Rule out the environment and the measurement before you blame hardware, because a dirty irradiance sensor can make every string look faulty at once.
Why is string 7 low?
A roof array has 12 strings of 20 modules each, and monitoring shows string 7 below the other eleven on every sunny day. The numbers are made up for the example, and the diode behaviour is simplified.
Is it the sensor? All twelve strings share one irradiance sensor, so a sensor fault would hit them equally. One string alone points at that string. Does the gap grow with the sun, or sit there from first light? A gap at one time of day suggests shade. Monitoring and performance data is the place to start.
Is anything visible? From the ground or another safe position, check for broken glass, scorching, a loose cable or a new shadow.
Does it need a test? If nothing shows, ask an authorised tester to measure string voltage and current under a written method, with every string measured within the same few minutes by the same instrument.
What does the result say? String 7’s open-circuit voltage comes back about 1.7% below its neighbours. A bypass diode that failed short removes its group of cells, often about a third of a module: 1 / (3 x 20) = 1/60, about 1.7% of the string voltage. That fits, though temperature also moves voltage, so only the neighbour comparison counts. A stepped I-V curve (Fluke’s guide lists shorted bypass diodes among the causes of steps) or a thermal image taken in good conditions could support it.
The outcome is a warranty claim for one module, with readings, conditions and dates attached, not a new string. Inverter faults, repairs and warranty claims shows what a claim needs.
Live DC work is for trained, authorised people
A module in daylight is a live source, and the Energy Safe Victoria and WorkSafe Victoria guideline warns that solar DC cables may still be live with the power off. HSE’s GS38 expects work to be done dead wherever possible. Many measurements above are live DC work for trained people following a written method. This page teaches the reasoning, not the work: stay within your local licensing rules and do not try a test from reading alone. Test instruments and safe testing covers isolation and test leads.
Write each reading down with its conditions
A number without conditions cannot be repeated or used in a warranty case. Record the date and time, plane-of-array irradiance, module temperature, string, instrument and a photo of the fault. These slips spoil a record:
- Short-circuit current with no irradiance logged: the number cannot be compared, so log irradiance at the same moment, or discard the reading.
- Treating one reading as proof: repeat it, or confirm it with a second kind of evidence.
- Replacing only the visible damage: a melted connector got hot for a reason, so ask what caused it before you fit the new one.
What you will learn
- A fixed order of diagnosis: data first, then a visual check from a safe place, then authorised tests.
- What open-circuit voltage, short-circuit current, insulation resistance, I-V curves and infrared images each tell you.
- How bypass diode failures, hot spots, connector damage and potential-induced degradation show up.
- How to read the six common I-V curve deviations and rule out shade, dirt and measurement error first.
- How to record readings with their conditions so another technician can repeat them.
The parts of this topic
A diagnostic method: data, look, then testabout 2 hours
Start with the least intrusive evidence and move inward, so you do not miss a simple cause or take an unnecessary risk.
Monitoring shows which string, since when and whether the shortfall follows the sun. A visual check from a safe position looks for broken glass, scorching and loose cable. Only then do authorised tests begin. The IEA PVPS review of failure algorithms lists four routes to evidence: electrical signatures, history versus today, predicted versus actual, and comparison with similar subsystems. A dirty irradiance sensor, for instance, can explain a whole plant's apparent shortfall, so check it before reaching for the most exciting instrument.
Open-circuit voltage and short-circuit currentabout 2 hours
Voc follows cell count and temperature; Isc follows sunlight. Comparing strings against neighbours shows which one is off.
Open-circuit voltage depends mainly on the number of cells in series and on temperature; short-circuit current depends mainly on sunlight. Compare each string with its neighbours, corrected for conditions. A low Voc often means part of a string is bypassed or missing; a low Isc often means shade, dirt or a poor connection. These are live DC measurements, made by trained and authorised people to a written method, such as the IEC 62446-1 test regime. Readings taken minutes apart under moving cloud cannot be compared fairly.
Insulation resistance and earth faultsabout 2 hours
A low insulation reading means current can leak to earth: a shock and fire concern, and a common reason an inverter refuses to start.
Insulation resistance testing checks that current cannot leak from live conductors to earth. Low readings point to a damaged cable, water in a connector or junction box, or a cracked module back, and inverters often refuse to start with an insulation alarm. The Victorian safety guideline warns that solar DC supply cables may still be live with the power off, because light keeps the modules energised, so the method must control the source. Example: one string far lower than its neighbours after rain. Do not pull connectors apart in wet weather to look.
- PV Installers Safety Guideline Energy Safe Victoria and WorkSafe Victoria · Guide Victoria, Australia
Connector failures and cable damageabout 2 hours
Connectors are a common place for heat, arcing and fire, and wrong tools or mixed parts are a recurring cause.
The UK MCS standard notes that the wrong tools for crimping and assembling DC connectors are a significant contributor to connection failure and even fires. It also asks that inverter arc-fault detection, where fitted, is enabled, and that drip loops and bottom cable entries keep water out. Swapping the visible damage without asking why it failed (tool, mixed parts or water) invites a repeat. Look for discolouration, melted housings, mixed brands, and chewed or abraded cable: a string can keep running with a melted connector in it.
- MIS 3002: the Solar PV Installation Standard (2025, issue 2.0) MCS · Standard UK, up to 50 kWp DC; free for personal use per MCS; check the library for the current issue
Bypass diodes and hot spotsabout 2 hours
Bypass diodes let current skip a shaded or damaged group of cells; when they fail, output drops or a cell overheats.
Each module has bypass diodes across groups of cells. A diode failed open can leave a shaded or damaged cell to overheat; a diode failed short permanently removes its group, often about a third of the module. A diode is not something to test in place: the string has to be isolated first, by someone authorised. Hot spots also come from cracked cells, poor solder joints and long-term shading. The IEA PVPS report on failure modes names junction-box faults, including faulty bypass diode connections, and recommends 100% testing during production and in affected installations.
- Degradation and failure modes in new PV cell and module technologies IEA PVPS Task 13 · Guide Free PDF
Potential-induced degradation and ageingabout 1 hour
Potential-induced degradation is a power loss tied to the voltage between cells and earth, helped by heat and humidity.
PID reduces module power when the cells sit at a high voltage relative to earth, and heat and humidity make it worse. In data it appears as a gradual loss on particular strings; the failure fact sheets say it depends on string voltage and polarity, and that some modules are designed to resist it. The IEA PVPS report says mitigation exists at cell, module and system level. Before replacing modules, ask the manufacturer what mitigation exists, and record dates and readings for any warranty case.
- Photovoltaic failure fact sheets 2025 IEA PVPS Task 13 · Guide Free PDF; annex to the failure-modes report
Reading an I-V curveabout 3 hours
Check the curve is valid, then match its shape to one of six deviations: steps, low Isc, low Voc, round knee, series resistance or shunting.
A trace plots current against voltage as the load moves from open circuit to short circuit. First check it is valid: loose leads cause dropouts and spikes. Then compare with a predicted curve for the same conditions. Rule out environment and measurement first, and do not blame ageing from one rounded knee. A stepped curve means bypass diodes are activating, from shade, dirt or damaged cells. Low Isc means less light or a wrong reading. A low voltage ratio at maximum power suggests series resistance; a low current ratio can point to shunting or a thin sliver of shade.
Thermography for diagnosisabout 2 hours
Infrared images show where heat builds up, which tells you where to look, not why it happened.
A thermal camera highlights a hot cell, connector or junction box under load. Conditions matter: IEC TS 62446-3 sets sunlight and wind limits, and Sitemark's article says the specification was written for handheld inspection, with parts that are impractical for drones. A warm cell in an image taken in weak light is not a fault, and a real one can be missed from the wrong angle. A hot spot shows where, not why: confirm with an I-V curve or visual evidence.
- Is Sitemark IEC compliant? Sitemark · Guide Vendor blog; useful for what the standard asks and where drones struggle
Practice task
Build a one-page paper decision chart: for each of six I-V curve symptoms, list the likely causes in order, what other evidence you would ask for and which causes are safety hazards. Then use pvlib's I-V curve example (free, needs Python) or a spreadsheet to draw a healthy curve and one with extra series resistance, and annotate the difference. This is analysis only; do not test any real circuit to complete it.
Skills you build
Fault diagnosis methodI-V curve interpretationInsulation and earth-fault reasoningThermography interpretationEvidence recording
Resources for this topic
Every source here is free. Each link opens the provider's own site in a new tab.
- GS38: electrical test equipment for use on low voltage systems HSE (United Kingdom) · Guide Free copy; UK guidance, its principles apply anywhere
- DC module I-V characteristics Sandia PV Performance Modeling Collaborative · Guide
- What is I-V curve tracing in solar? Fluke · Guide Vendor blog that also sells testers; the explanation is useful, ignore the product push
Latest Renewable Energy jobs
No title matches the role right now, so these are the latest jobs in the same category.
No live jobs to show right now. Check back soon, or browse every green job.
Spot a problem with this topic?
Tell us what is wrong, missing or unclear here. Your note is attached to this topic.
Notes are sent from this page with JavaScript on. Without it, write to us on the contact page.