PV Module Insulation Resistance, Wet Leakage Current and Dielectric Withstand: System-Voltage and RFQ Evidence

A PV module can produce the expected watts and still have an unacceptable electrical-insulation system. A flash report describes power at specified measurement conditions; it does not prove that the cell circuit remains isolated from the frame, glass edge, back surface, junction-box exterior or wet accessible surfaces. That distinction matters because a weak insulation path can create shock risk, ground faults, inverter insulation alarms, corrosion or intermittent field failures.

For procurement, the direct answer is simple: do not approve a module from a single “insulation passed” statement. Require the exact module model and construction to be mapped to a current safety-qualification report, then review the dry insulation/dielectric and wet leakage results with the rated maximum system voltage, application class, polarity, test sequence, specimen identity, equipment traceability and post-stress condition. Confirm that the report covers the production BOM and factory being purchased. Treat a certificate, a test report and a routine production test as three different evidence layers.

This guide explains how to request that evidence without copying an obsolete voltage, time or pass criterion into an RFQ. It complements the PV module glass and backsheet BOM change-control guide, the thermal-cycling and damp-heat guide and the flash-test and EL traceability guide. It does not certify any module, prescribe high-voltage laboratory work or claim an unverified SINAWATTS certification, laboratory, test capability, model, factory, material, limit, stock, price, MOQ or lead time.

Direct answer: what should a module RFQ require?

For each offered model, require the bidder to return:

• exact manufacturer, production site, model designation and nameplate maximum system voltage;
• safety class/application classification and the market-specific standard edition claimed;
• current certificate number, issuing certification body and scope or model-family appendix;
• complete accredited or otherwise buyer-approved test report, not a cropped pass page;
• report-form edition and deviations or national differences;
• dry insulation-resistance and/or dielectric-withstand test identification, as applicable;
• wet leakage-current or wet insulation test identification, as applicable;
• test voltage, polarity, ramp, dwell, measured quantity and pass criterion copied from the applicable controlled method or report;
• specimen dimensions and any area-based normalization used by that method;
• exposed-conductive-part, frame, foil, bath or wetting arrangement;
• preconditioning and sequence position for every reported measurement;
• initial and post-stress actual results, not only “P”;
• test equipment ID, calibration status, measurement range and uncertainty/decision-rule statement where relevant;
• BOM identity for glass, encapsulant, cells, interconnects, backsheet or rear glass, edge seal, frame, junction box, potting, cables and connectors;
• model-family mapping showing why the tested construction covers the quoted model;
• retest assessment for every difference between the tested and offered construction;
• production dielectric/insulation test plan, limit, frequency and traceability; and
• signed exceptions where any requested evidence is unavailable.

The buyer should place the required system voltage and installation environment in the RFQ. The supplier should identify the applicable test path. A buyer should not invent a test recipe from a memory of an older IEC or UL edition.

Keep four different questions separate

“Insulation test” is often used for several different decisions. Separate them before comparing bids.

On small screens, swipe or scroll sideways to read every column. Keyboard users can focus the table and use the arrow keys.

A fifth question is reliability after stress. A module may pass initially but fail after humidity, temperature cycling, mechanical loading or another sequence. That result is different from an initial safety check and should be reviewed as a separate row.

Use current standards and editions, not a remembered test number

The official IEC page for IEC 61730-2:2023, checked on 2026-09-30, says the document lists tests required for PV-module safety qualification and is used with IEC 61730-1. The stated objective is to verify the safety of a construction assessed under Part 1, with sequences and criteria intended to detect breakdown that could lead to fire, electric shock or personal injury. The official IEC 61730-1:2023 page, also checked on 2026-09-30, describes the construction requirements and notes that modified constructions are qualified through IEC TS 62915.

The official page for IEC 61215-2:2021, checked the same day, describes design-qualification test procedures intended to determine electrical characteristics and demonstrate, within practical constraints, resistance to prolonged outdoor exposure. IEC 61215 and IEC 61730 have related tests and sequences, but their purposes are not interchangeable: design qualification is not a substitute for safety qualification, and a safety certificate is not a lifetime guarantee.

The edition matters. Test names, sequence logic, construction requirements, report forms and national deviations can change. Ask the bidder to state the exact edition and amendment/corrigendum basis. If a certificate cites an older edition, request the certification body’s status and the buyer’s market-acceptance decision rather than declaring it valid or invalid from the date alone.

Start with maximum system voltage and the complete application boundary

Electrical stress is not chosen from module wattage. The RFQ should state at least:

• maximum DC system voltage;
• grounded, functionally earthed or floating-array architecture;
• expected polarity of the cell circuit relative to earth over operating modes;
• installation class and accessible conductive parts;
• frame material and intended bonding/grounding method;
• rooftop, ground-mount, mobile or other use;
• wet, coastal, agricultural or high-condensation exposure;
• altitude if relevant to insulation coordination;
• module operating-temperature envelope;
• compatible cables, connectors and junction-box arrangement; and
• any national code or certification route required by the destination market.

Do not assume a nameplate value automatically proves that every insulation component has matching evidence. The test report should connect the system-voltage rating to the evaluated construction. The rear dielectric stack, glass edge, frame bond, junction-box attachment, potting, cable entry, lead insulation and connector interfaces all contribute to the boundary.

A frame-less glass/glass module and a framed glass/backsheet module create different accessible surfaces and test setups. A conductive backsheet or metal insert changes the boundary again. The report should describe the actual specimen rather than using only a commercial family name.

Understand dry dielectric withstand, insulation resistance and wet leakage

These terms answer related but distinct questions.

Dielectric withstand

A dielectric-withstand test applies an electrical stress between the active circuit and accessible conductive parts or a defined external electrode. The observation may include breakdown, flashover or leakage within the method’s criteria. The voltage, waveform, polarity, ramp and duration belong to the applicable standard/report. A high-voltage value copied without those conditions is incomplete.

Insulation resistance

Insulation resistance is a measured resistance between defined conductive regions at a stated DC test voltage and time. Readings can be affected by module area, surface condition, humidity, temperature, fixture leakage, stabilization time and instrument range. “Greater than meter limit” is meaningful only when the range and setup are recorded.

Wet leakage current or wet insulation

The module is deliberately exposed to a controlled wetting medium so that edges and surfaces that may form a leakage path are challenged. Wetting chemistry, water condition, temperature, coverage, connection arrangement and test sequence matter. A water spray photograph is not a report. A dry megohmmeter result cannot replace wet evidence.

A historical NREL/SERI report, TR-213-3624, explains the engineering purpose of wet insulation testing: to evaluate whether moisture associated with rain, fog, dew or melted snow can reach active portions and create corrosion, ground-fault or shock hazards. It also shows why old public procedures must not be copied into a current purchase specification. That interim thin-film document is historical research, not the current IEC 61730-2:2023 acceptance method. Use it to understand failure physics, never as a substitute for the applicable current standard.

Require the actual test configuration

A useful report should let a qualified reviewer reconstruct the electrical boundary. Request:

• how positive and negative output leads were joined or energized;
• polarity or polarities applied;
• where the return electrode was connected;
• whether a metal frame, grounding point, conductive foil, bath or plate was used;
• which module faces and edges were wetted;
• how the junction box, cable entry and leads were treated;
• module and liquid temperature;
• wetting solution identity and controlled properties where the method requires them;
• voltage ramp and dwell;
• measurement timing;
• instrument and fixture leakage checks;
• discharge and safety controls; and
• the exact criterion and recorded result.

Photographs can help confirm fixture geometry but do not replace the controlled data sheet. The report should distinguish measured module leakage from test-lead, bath, fixture or instrument leakage. If a value is corrected or normalized, the raw value, formula and inputs should remain traceable.

Do not ask a production operator to reproduce a qualification-laboratory wet test unless the safety method, equipment, competence and process are formally established. High-voltage and wet testing require controlled laboratory procedures. Procurement’s role is to request evidence and a safe production-control plan, not to improvise the test.

Compare results only on the same basis

Two megohm values cannot be compared until the basis is aligned. Review these fields:

On small screens, swipe or scroll sideways to read every column. Keyboard users can focus the table and use the arrow keys.

For example, “>2 GΩ” and “1.6 GΩ” do not prove the first construction is better. The first may be limited by the meter range; the module areas, test voltage and moisture condition may differ. The relevant decision is whether each result meets its applicable criterion with valid model coverage and test integrity.

Map every specimen to the production BOM

Insulation performance can depend on interfaces that a sales datasheet does not show. The report-to-BOM map should include:

• front glass type, thickness and coating where relevant;
• encapsulant manufacturer, grade and thickness;
• cell technology and layout;
• ribbons, busbars and edge distances;
• rear glass or backsheet manufacturer, grade, layer construction and thickness;
• edge seal or perimeter construction;
• frame alloy, finish, geometry and grounding feature;
• adhesives and tapes at the frame and junction box;
• junction-box body, lid, potting, adhesive and mounting footprint;
• bypass-diode and internal conductor arrangement;
• output cable type, insulation, length and entry seal;
• connector family, contact, seal and assembly; and
• manufacturing site and critical process controls.

The module-lead and connector BOM guide explains why a cable or connector substitution cannot disappear inside a generic model number. The junction-box and bypass-diode guide covers the termination assembly. Both interfaces sit on the insulation boundary.

A certificate may list a broad model family. The detailed report and construction data determine which variants were actually represented. Ask for a matrix with tested worst cases and the rule used to cover other sizes, powers, cell counts or rear constructions.

Treat changes as a retest decision

The official IEC TS 62915:2023 page, checked on 2026-09-30, describes a uniform approach for maintaining design and safety qualification when PV modules are modified. It states that the 2023 edition separates IEC 61215 and IEC 61730 retest requirements and includes a matrix relating modifications to retesting. It also references component-level IEC 62788 material standards.

That means “same wattage” or “same datasheet” is not a change-control decision. Require a documented assessment after changes to:

• backsheet or rear glass;
• encapsulant or cure process;
• glass coating or thickness;
• cell or interconnect layout affecting edge distance;
• frame geometry, finish or grounding provision;
• junction-box body, potting or adhesive;
• cable entry, output cable or connector;
• edge seal;
• module dimensions;
• maximum system voltage;
• factory or lamination process; or
• model-family naming/coverage.

The official IEC TS 62788-2:2024 page, checked on 2026-09-30, describes methods and reporting for mechanical, electrical, thermal, optical and chemical properties of polymeric frontsheets and backsheets in representative combinations with matched materials. Material data can support change review, but component data alone do not prove the finished module passes its required module-level tests.

Review initial and post-stress insulation evidence

An initial pass is only one point. Review where insulation or wet leakage checks occur in the qualification sequence and which stress preceded each check. Depending on the applicable method and program, the evidence package may contain results associated with humidity, thermal cycling, mechanical loading, UV or other conditioning.

Use a sequence table:

On small screens, swipe or scroll sideways to read every column. Keyboard users can focus the table and use the arrow keys.

The table above is a procurement template, not a prescribed IEC specimen allocation. Populate it from the current report.

The mechanical-load testing guide shows why mounting method and load history matter. The thermal-cycling and damp-heat guide explains how thermal and moisture stresses interact with the BOM. Do not detach an insulation result from the sequence that produced it.

NREL’s 2013 technical report Photovoltaic Module Qualification Plus Testing, NREL/TP-5200-60950, used insulation resistance and wet leakage as repeated checks around extended stresses and required test samples to be selected from shipping production. That document is research guidance, not a current universal purchase requirement. Its useful procurement lesson is that specimen selection and before/after checks help reveal whether stress has weakened the insulation system.

Diagnose weak or inconsistent results without inventing a cause

A low or unstable reading is evidence of a problem, not proof of one specific cause. Possible paths include:

• moisture at a glass/backsheet or glass/glass edge;
• backsheet pinhole, crack, scratch or fold damage;
• insufficient encapsulant distance around live parts;
• junction-box potting void or adhesive discontinuity;
• cable-entry seal leakage;
• damaged lead insulation;
• conductive contamination;
• frame contact with an unintended conductor;
• test fixture leakage;
• wetting inconsistency;
• incomplete stabilization or capacitive charging; or
• instrument range or connection error.

Require containment of the affected lot and a documented investigation. The supplier should preserve the failed specimen and original data, confirm the test system with a known reference, repeat only under an approved troubleshooting plan, localize the path where possible, inspect the construction and compare sibling samples. An uncontrolled retest-until-pass rule destroys evidence.

Trend actual results by model, BOM revision, line, shift and stress condition. A downward trend inside the pass range may identify process drift. Trend analysis does not justify a new acceptance limit; it informs investigation and preventive control.

Distinguish qualification, surveillance and routine production tests

Use three evidence layers:

1. Qualification: representative specimens assessed under the named design and safety standards.
2. Certification maintenance or surveillance: evidence that the listed construction and factory controls remain under the certification scheme.
3. Production control: routine or sampled tests used to detect manufacturing defects.

The official IEC listing for IECEE TRF 61730-2F:2024, checked on 2026-09-30, says that test-report form applies to IEC 61730-2:2023 together with IEC 61730-1:2023. The IEC webstore FAQ also states that a test report form is not a valid CB Test Report unless it is signed by an approved CB Testing Laboratory and appended to a CB Test Certificate issued by a National Certification Body. Therefore, a blank TRF, an unsigned spreadsheet or a supplier-created page is not equivalent to a certificate-backed CB report.

For routine production evidence, request:

• controlled procedure and revision;
• test boundary and method;
• voltage/current/resistance limits as applicable;
• instrument make/model/ID and calibration status;
• fixture verification and leakage check;
• operator authorization;
• serialized result linked to the module label;
• automatic recording or protected manual entry;
• fail lockout and retest disposition;
• reaction plan for equipment drift; and
• record-retention period.

Do not assume a production test uses the same stress or purpose as qualification. Ask the supplier to explain the relationship.

Verify certificates and reports as a joined evidence set

A credible packet should let the buyer answer five questions:

1. Authenticity: Can the certificate number and issuer be verified through the certification body or official scheme route?
2. Currency: What edition, amendment, issue date and status apply?
3. Scope: Is the offered model/family and production site included?
4. Construction: Does the detailed report identify the actual tested BOM?
5. Continuity: Does change control show that the production BOM still matches or has an approved retest assessment?

A logo on a datasheet answers none of these alone. Ask for the certificate, model appendix, report and construction mapping together. If confidential details are redacted, require enough controlled information for an authorized reviewer to confirm coverage. Record unresolved redactions as an open item rather than guessing.

Use a bounded comparison example

Assume an RFQ requests framed modules for a 1,500 V DC project in a humid site. The values and bid descriptions below are invented to demonstrate method; they are not acceptance criteria or product claims.

Bid A provides a marketing certificate image and says “insulation >1 GΩ.” It does not identify wet or dry method, test voltage, module area, standard edition, specimen model, production site or post-stress condition. The value cannot be compared or accepted.

Bid B provides a complete older-edition report for a 1,000 V family. The offered datasheet now says 1,500 V, and the backsheet plus junction box have changed. The supplier gives no retest decision. Bid B may be a useful historical baseline, but it does not demonstrate the offered 1,500 V construction.

Bid C provides a current certificate and signed report mapped to the exact glass/backsheet, encapsulant, frame, junction box, cable and connector. It states the maximum system voltage, lists initial and post-stress dry and wet results, identifies the current report-form basis, documents the factory and supplies the retest assessment for a cable-length variant. It also returns the serialized production insulation-test record format. Bid C is the strongest evidence package. Final acceptance still depends on the project’s qualified safety and certification review.

This comparison does not rank absolute resistance values. It ranks evidence completeness and applicability.

RFQ evidence matrix

On small screens, swipe or scroll sideways to read every column. Keyboard users can focus the table and use the arrow keys.

Keep “not demonstrated” as an allowed answer. It is safer and more actionable than converting missing evidence into an assumed pass.

Put an evidence gate before purchase release

A practical gate can be five steps:

1. Identity gate: model, site, label and BOM are unambiguous.
2. Certification gate: certificate and report are authentic, current enough for the project and in scope.
3. Test gate: dry, dielectric and wet evidence is applicable, traceable and complete.
4. Change gate: offered construction differences have documented retest decisions.
5. Production gate: routine controls link shipped serial numbers to controlled test records.

Close every open item with a named document and responsible approver. Do not use a supplier’s promise to send evidence after shipment as closure.

Source and evidence boundaries checked on 2026-09-30

The IEC pages for IEC 61730-1:2023, IEC 61730-2:2023, IEC 61215-2:2021, IEC TS 62915:2023, IEC TS 62788-2:2024 and IECEE TRF 61730-2F:2024 were checked on 2026-09-30. They establish scope, current editions and the role of construction, testing, retesting and report forms. The public pages do not reproduce every controlled test voltage, duration, calculation or pass criterion. Those details must come from the licensed standard and the signed report used for the transaction.

The two linked NREL documents provide historical and research context for wet insulation and repeated post-stress checks. They do not supersede current standards or define acceptance for every module. A supplier’s certificate and results apply only to the models, construction, sites, conditions and revisions actually covered. National requirements and project approval remain with the responsible certification body, laboratory, designer and authority having jurisdiction.

Send the module model list, system voltage, destination market, installation environment and available certificate/report package for a structured RFQ review. Ask bidders to return the evidence matrix with exact model and BOM references. The actual supplier must confirm certification scope, production site, availability, price, quantity and lead time.

Buyer FAQ

Is an IEC 61215 report enough to prove module electrical safety?

No. IEC 61215 addresses design qualification and type approval for outdoor performance/reliability objectives. Electrical safety qualification is addressed through the applicable IEC 61730 construction and test requirements, plus national or regional requirements. Review both where the project requires them.

Does a flash-test report include insulation evidence?

Normally it addresses electrical power characteristics at specified measurement conditions. It is not a substitute for insulation, dielectric-withstand or wet leakage evidence. Link the flash report and safety report through the same serial/model/BOM traceability.

Is a certificate page enough?

No. A certificate helps establish scheme status and scope, while the detailed report shows specimens, construction, methods and results. The buyer also needs change-control evidence tying current production to the tested construction.

Can I compare two suppliers by the highest megohm value?

Not without aligning test method, voltage, area, temperature, moisture, timing, instrument range and sequence position. Evidence completeness and applicable pass margin matter more than an isolated headline value.

Why does module area matter?

Larger surfaces and edge lengths can affect leakage opportunity, and some methods use area-related criteria or normalization. Use the rule in the applicable current method and record the actual dimensions; do not invent a universal conversion.

Is a dry dielectric-withstand pass equivalent to a wet leakage pass?

No. The wet test deliberately challenges surface and edge leakage paths under controlled wetting. Each test has its own setup and acceptance purpose.

What if the measured value is above the instrument range?

Record it as a censored result such as “greater than the verified range,” together with the range and method. Do not turn it into an invented numerical value or use it to rank suppliers beyond what the instrument demonstrated.

Should every production module undergo the same wet test as qualification samples?

Not automatically. Qualification and routine production tests have different purposes and safety controls. Require the supplier’s approved production test plan and the applicable certification requirements; do not improvise wet high-voltage testing on a production line.

What changes should reopen the evidence review?

Changes to rear insulation, encapsulant, edge seal, frame, junction box, potting, adhesive, cables, connectors, module dimensions, live-part spacing, system voltage, factory or critical process should trigger a documented assessment. The applicable retest matrix determines what additional testing is required.

Can component certificates replace a module report?

No. Component evidence can support material selection and change assessment, but the finished module’s interfaces and construction require module-level evaluation.

What should happen after an insulation failure?

Contain affected material, preserve the failed specimen and original record, verify the test system, investigate the leakage path, assess sibling products and document corrective action. Do not erase the failure through uncontrolled repeat testing.

Does passing qualification guarantee 25 or 30 years of field life?

No. Qualification applies defined stresses and pass criteria within practical constraints. It supports design and safety assessment; it is not a universal lifetime prediction or warranty validation.