A solar simulator class describes specified characteristics of the illumination; it does not, by itself, establish the uncertainty of a module's reported power. Ask for two linked records: the classification report covering the actual measurement configuration, and the laboratory's uncertainty statement for the reported electrical quantity. Both must apply to the module and test conditions being purchased.
For a useful RFQ, identify the module dimensions and technology, required reporting condition, purpose of the measurement, simulator configuration, usable test area and required result package. Then ask the provider to explain how its class claim and uncertainty evaluation support that particular job. A brochure badge alone cannot close those questions.
This guide helps purchasing teams review laboratory and production-test evidence. It does not prescribe a calibration procedure or claim that SINAWATTS owns a particular simulator, operates an accredited laboratory or can provide every measurement discussed. Available documentation and any third-party work require confirmation for the proposed order.
Start with the decision the measurement must support
A buyer may need routine shipment records, a comparison between two module designs, an independent check near a contractual power limit, or a reference measurement for another test system. These decisions do not necessarily require identical deliverables. State the decision before asking suppliers to quote a more impressive class label.
For routine production evidence, the immediate question may be whether the results are controlled and traceable to the delivered units. For a narrow performance comparison, the question also becomes whether the measurement process can support the distinction being claimed. A provider should know which interpretation the customer intends to make.
Write one sentence in the request: "The measurement will be used to decide whether the identified module meets requirement P-04 under the agreed reporting conditions." Attach the requirement, define the quantity and identify the reviewer. This prevents a laboratory from supplying a technically competent report that answers a different question.
Keep serial-to-report matching in the flash-test and EL traceability workflow. Keep response curves and device-specific spectral adjustments in the spectral-mismatch evidence guide. The present review concerns the coverage of the simulator classification and its relationship to the uncertainty of the complete measurement.
Read the three classification fields and the edition
The official IEC 60904-9:2020 publication record identifies three classification criteria: spectral distribution match, irradiance non-uniformity in the test plane, and temporal instability. It lists A+, A, B and C classes and describes the 2020 document as edition 3.0. The public catalog also notes the addition of an extended spectral range and the A+ class in that edition.
That public description establishes scope, not the full test procedure or compliance of an offered machine. Request the classification report, the cited edition and the result for each criterion. Where a shortened label is used, have the provider expand it into named fields so nobody has to infer the order or meaning of the letters.
Do not silently compare a legacy certificate with a report under a different edition. Ask the provider to describe the methodological difference relevant to the proposed measurement and to state the accepted basis. An older report is not automatically false, but its scope must not be rewritten as though it were issued under a later document.
Pasan's manufacturer explanation of simulator classes distinguishes its A++ terminology from the classes published in the 2020 IEC edition. The purchasing lesson is to request the numerical definition and method behind any additional vendor designation. Do not treat a promotional extension as a universal IEC grade shared by every manufacturer.
Separate the classification claim from the result claim
Use two evidence columns in the bid comparison. The first records what the illumination report establishes. The second records what the laboratory says about the uncertainty and applicability of the actual electrical result. Leave a gap visible when a supplier answers only one column.
On small screens, swipe or scroll sideways to read every column. Keyboard users can focus the table and use the arrow keys.
| Evidence question | Record to request | What the record does not establish alone |
|---|---|---|
| Which classification is claimed? | Named criteria, edition, results and report identifier | Uncertainty of every measurement performed on the machine |
| What area and configuration were evaluated? | Test-plane dimensions, positioning and operating configuration | Coverage of a larger or differently positioned module |
| Which result is being reported? | Quantity, units, conditions, sample identity and released report | Suitability of another quantity or a different test method |
| How is uncertainty stated? | Quantity-specific statement, evaluation basis and coverage information | Automatic compliance with the customer's acceptance rule |
| Why does the evidence apply here? | A linked applicability statement for the proposed work | Approval of unlisted variants or later configuration changes |
This is an original purchasing worksheet, not an extract from a standard. Its purpose is to stop an attractive answer in one row from replacing an unanswered question elsewhere. The technical reviewer can add rows for the particular measurement without changing that separation.
As a concrete research example, a 2023 Acta IMEKO study describes a class AAA pulsed system and separately reports uncertainty estimates for current, voltage, maximum power and efficiency. Its abstract gives a relative expanded combined uncertainty of 2.05% for maximum power with coverage factor k = 2. This belongs to that study's system and measurements; it is not a general AAA specification or a benchmark promised for other laboratories.
The example demonstrates why a class label and a measurement uncertainty belong in separate fields. It does not justify substituting the paper's number for a supplier's own evaluation. Request the actual laboratory's evidence for the actual task.
Match the classified test area to the module footprint
Ask the provider to show the assessed test plane and the intended position of the module on the same drawing. Record the dimensions, orientation and placement references. A statement that a simulator can physically accommodate a module should be distinguished from evidence that the relevant measurement area is covered by the claimed classification.
For an unfamiliar module size, ask how the provider establishes coverage of the active device in the proposed position. Do not assume the center of a large illuminated field represents its edges. Equally, do not invent an exclusion zone around every test area; obtain the applicable report and the provider's interpretation.
The RFQ should identify the sample outline and any fixture that changes its position. A thicker frame, alternative support or different clamping arrangement may move the device relative to the declared plane. Have the provider state whether its accepted configuration covers that arrangement before transferring an existing approval.
Make dimensional exceptions specific. "Large module supported" is too broad for purchasing release. A better response identifies the drawing revision, evaluated area, proposed module orientation and any limitation requiring technical review. Preserve that response with the report so the production team receives the same arrangement that was approved during sourcing.
Check operating configuration and measurement timing
Request the configuration used for classification and the configuration proposed for the module measurement. Useful identifiers include the light-source system, relevant settings, optical arrangement, test-plane location and acquisition method. The report owner should say which details affect applicability rather than leaving a purchaser to reconstruct the apparatus.
For a pulsed system, ask how the measurement acquisition interval relates to the illumination interval represented by the evidence. A pulse-duration headline does not describe every part of the acquisition process. Obtain a provider explanation appropriate to the module technology and reported quantity, including the basis for choosing the measurement method.
For a steady-state system, request the applicable operating and stabilization conditions. Avoid imposing a generic waiting time copied from another machine. The laboratory's controlled method should explain when a result is eligible for reporting and what records demonstrate that the required conditions were met.
Treat maintenance and configuration changes as applicability questions. Ask which changes require review of the classification or uncertainty records and who authorizes continued use. The buyer does not need to set a universal lamp lifetime or recalibration interval; it needs a documented control process and an identifiable status for the proposed measurement date.
Ask for uncertainty in the quantity you will compare
A percentage labeled "accuracy" is incomplete if the reported quantity, reference basis and interpretation are missing. Ask whether the number relates to irradiance, current, voltage, maximum power or another quantity. Also ask whether it describes an instrument specification, repeatability observation, standard uncertainty or expanded uncertainty.
NIST Technical Note 1297's expanded-uncertainty guidance describes expanded uncertainty as a coverage factor applied to combined standard uncertainty and calls for reporting the coverage factor used. This is general metrology guidance, not a PV acceptance standard. It supports asking the provider to explain its statement rather than accepting an unexplained plus-or-minus percentage.
For a relative uncertainty, retain the quantity used as the denominator and the corresponding absolute value where helpful. A percent of measured power is different from a percent of another reference value. If offers use different conventions, ask each provider to restate the results on the agreed reporting basis before commercial ranking.
Request an uncertainty summary with the report identifier and applicability conditions. Where a detailed budget is commercially restricted, agree a review arrangement that still lets the responsible specialist assess the claim. A confidentiality limitation should be documented; it should not be converted into an assumption that the missing evaluation must be satisfactory.
Review a budget without pretending to audit the whole laboratory
The purchasing team can check that the uncertainty summary names its major contributing measurement elements and points to controlled supporting records. The competent measurement specialist should assess the evaluation itself. This division keeps the buyer's completeness review useful without presenting it as a substitute for technical assessment.
Ask how the provider treats the reference device, electrical measurements, temperature information, illumination effects and data processing insofar as they apply to the method. The response should say which terms are included, which are judged negligible with justification and which are outside the stated scope. This is an evidence request, not a universal list of mandatory terms.
NIST's photovoltaic characterization laboratory description illustrates the distinct roles of spectral-responsivity measurement, irradiance calibration and indoor current-voltage measurements. It also describes spectral-mismatch correction in its own measurement process. Those are NIST capabilities, not evidence that a supplier uses the same instruments or achieves the same results.
For an outsourced measurement, identify the entity responsible for the released result. A module supplier may forward a laboratory report without owning the method or uncertainty budget. Direct technical questions to the report owner through the agreed channel and keep the supplier's commercial responsibilities distinct from the laboratory's measurement statements.
Send the module drawing, required measurement decision and available simulator report with your RFQ. State which evidence is missing and request written confirmation of the documents or third-party scope available for the proposed supply.
Compare two hypothetical reports without inventing a winner
The following figures are invented for a document-review exercise. They are not product ratings, laboratory performance claims or customer results. Suppose Report A states 510 W with relative expanded uncertainty of 2.0%, while Report B states 518 W with 1.5%. Both identify k = 2, and both use the same agreed reporting conditions.
The corresponding stated uncertainty magnitudes are 10.2 W and 7.77 W. The nominal difference between the reported powers is 8 W. These calculations help a buyer read the reports, but they do not establish a formal comparison result. The reviewer still needs the evaluation basis, relevant correlations, sample equivalence and decision rule.
Do not award the higher number simply because both reports mention class AAA. Do not declare the two modules equivalent merely because simple plus-or-minus intervals overlap. Either conclusion would skip the actual comparison method. Ask the responsible specialist what conclusion the available evidence supports and what further work, if any, is justified.
On small screens, swipe or scroll sideways to read every column. Keyboard users can focus the table and use the arrow keys.
| Hypothetical submission | Useful information | Action before drawing a power conclusion |
|---|---|---|
| A: 510 W, 2.0%, k = 2 | A quantity-specific uncertainty statement | Confirm sample, conditions, budget scope and decision rule |
| B: 518 W, 1.5%, k = 2 | A different result with an identified statement | Align the measurement basis and assess comparison assumptions |
| C: 520 W, "AAA accuracy" | A nominal result and class claim | Request a separate uncertainty statement and applicability evidence |
The worksheet is deliberately not a pass/fail calculator. It helps procurement identify which comparison is ready for technical review and which remains incomplete. The outcome might be acceptance, a qualified comparison or a targeted additional measurement; that outcome must come from the agreed assessment.
Define the acceptance rule before results arrive
A contract may contain a lower power requirement, a permitted reporting basis or a rule for evaluating borderline results. Ask the buyer's technical owner to define how uncertainty will be considered before commissioning the measurement. The laboratory should acknowledge the requested decision rule or clearly state that it will report measurements without a conformity decision.
Preserve the distinction between a measured value, a conformity statement and a purchasing decision. They can appear in the same project file but are not interchangeable. A laboratory may issue a result that the customer then evaluates under a contract; another engagement may explicitly include a conformity assessment within the laboratory's agreed scope.
If the first result is borderline, follow the pre-agreed process. Do not change the rule, select the most favorable repeat or substitute a different reporting basis after seeing the number. If additional measurements are warranted, retain their purpose, sample identity and relationship to the earlier result.
For a disagreement between laboratories, request a structured reconciliation of sample condition, method, reporting basis and uncertainty evaluation. Start by identifying differences that could explain why the results are not directly comparable. A more expensive machine or a stronger class label is not, on its own, an explanation of the discrepancy.
Build a release record that survives changes and handover
A compact release record should connect the module drawing, measurement request, classification report, method revision, uncertainty statement and final report. Add the responsible technical review and any restrictions. This lets purchasing, quality and the receiving team use the same evidence without treating an email summary as the controlling document.
Consider an invented change request: the supplier proposes a longer module under the same commercial series name. The previous approval covered a specified footprint and fixture. The buyer should ask whether the new dimensions remain within the documented area and configuration and whether the measurement evaluation still applies. Matching the series name does not answer that question.
A change may be resolved with a documented applicability statement, a revised setup or additional evidence. Avoid demanding complete repetition for every administrative correction, but preserve the distinction between a spelling correction and a changed measurement arrangement. Record what was reviewed, by whom and against which version.
At shipment, match the delivered model and report package to the approved scope. Use the solar accessories overview for related sourcing context and the module drawing and dimensional inspection guide only if its drawing requirements are relevant to the order. Product navigation is not proof of a particular measurement service or certification.
Turn an applicability question into a closed record
Consider a fictional submission with classification report CL-08, fixture drawing FX-02 and module drawing MD-11. The supplier's cover sheet lists all three, but the proposed setup uses fixture FX-03. These identifiers are invented to demonstrate a review record, not to describe an actual laboratory or product.
A weak disposition would say "same machine, accepted." A useful issue entry says: "CL-08 references FX-02; the quotation proposes FX-03. The provider must identify the changed positioning features and state whether the classified plane, area and operating configuration remain applicable." The technical owner then reviews the response against the actual proposed measurement.
The closing evidence might be an accepted equivalence statement or a revised report; it should not be predetermined. If FX-03 is only a drawing-number correction with no physical change, document that fact. If it changes the device position, preserve the technical basis for continued coverage or the additional work required. Purchasing should not characterize the change from the file name alone.
Finally, link the disposition to the uncertainty statement and released measurement request. This prevents a closed classification question from being mistaken for approval of every other part of the measurement. The completed record tells the next reviewer precisely what changed, what evidence was considered and which decision was authorized.
Make the commercial request specific enough to price
Separate routine production records, independent measurements, technical interpretation and additional investigation into defined deliverables. Ask which items are included in the quotation, which require external arrangements and which depend on receiving samples or controlled data. This avoids comparing a basic data export against a quotation that includes substantial independent work.
Specify file formats and delivery milestones alongside technical scope. A readable report, a numerical result table and a mapping to sample identifiers may be enough for one project; another may need controlled raw-data access. Ask for only the evidence necessary to support the stated decision and preserve any agreed limitations.
Do not insert assumed prices, sample quantities, lead times or availability into the purchase order. Obtain written commercial confirmation after the provider has reviewed the actual request. If an evidence gap remains, identify whether it blocks the measurement, the conformity decision or the purchasing release so the project owner can act on the correct dependency.
FAQ
Does class AAA mean a PV power measurement has a fixed accuracy?
No. A class statement addresses specified illumination characteristics under its reported conditions. The uncertainty of the electrical result needs its own evaluation for the device, method and quantity. Ask for a separate uncertainty statement and applicability record rather than converting a class label into an assumed power percentage.
Is an A++ label automatically an IEC 60904-9 grade?
Do not assume that it is. The cited 2020 IEC publication lists A+, A, B and C. Manufacturers may use additional designations with their own definitions. Request the cited edition, numerical criteria, test method and report, and keep any vendor extension explicitly identified when comparing quotations.
Can a report for a smaller test area cover a larger module?
Coverage needs an explicit technical basis. Ask the provider to overlay the module position and relevant device area on the documented test plane and explain applicability. Physical clearance alone is insufficient evidence, while an unsupported assumption that all larger modules fail is equally unhelpful. The actual report and configuration decide the next step.
Is repeatability the same as measurement uncertainty?
Treat a repeatability result as one described observation, not an automatic substitute for the complete uncertainty statement. Ask what quantity was repeated, under which conditions and how the observation is used in the provider's evaluation. The responsible specialist should explain what the final reported uncertainty includes beyond that repeated measurement exercise.
What if the laboratory will not release its complete uncertainty budget?
Ask for the quantity-specific statement, coverage information, applicability conditions and a suitable technical-review arrangement. Confidential details may be handled under agreed terms, but the evidence must still support the purchasing decision. Record any unresolved limitation rather than replacing it with the simulator brochure or another laboratory's published numbers.
Should every maintenance action trigger a new measurement campaign?
The provider's controlled change process should determine the required review. Ask which action occurred, what configuration or evidence it affects and how continued applicability was established. A routine administrative update and a changed optical arrangement deserve different treatment; this guide does not prescribe a universal retest interval or automatic acceptance.
What should be included when requesting price and timing?
Provide the module identity and drawing, measurement purpose, required quantities and reporting conditions, available evidence, requested deliverables and review responsibilities. Ask for separately confirmed commercial terms and dependencies. Neither a class label nor this guide establishes availability, a laboratory service, a fixed cost or a promised delivery date.
Send a simulator-evidence and uncertainty RFQ
Send the module reference, test-area drawing, classification claim and required uncertainty statement for RFQ review. Identify the purchasing decision the measurement must support, attach existing reports and list unresolved scope questions. Request a written response covering available evidence, exceptions and commercial terms before releasing the order.