Evidence first.
The exact model, always.
SolarApprovals is an independent equipment directory for US installers, designers, and buyers. We make public records easier to inspect; we do not certify equipment or issue installation permits.
What a CEC listing means here
“Listed in CEC” means the exact manufacturer and model configuration was present in the active California Energy Commission workbook at the time of our source check. This is an observation, not a claim of national approval, current code compliance, eligibility for every incentive, or compatibility with your installation.
Confirm local permitting and interconnection requirements with your authority having jurisdiction (AHJ) and utility. Review the applicable certifications, firmware, and installation instructions for your exact configuration.
Two dates, two different meanings
“Checked” is when we retrieved the source. The publisher’s update note is preserved separately. Neither date is treated as the original listing or certification date. Historical status changes begin when this directory first observes a model.
Statuses follow evidence
- Listed in CEC: exact identity found in the active source.
- Not observed: a previously tracked identity is absent from a subsequent validated, complete workbook. This is not an inference that approval was revoked. Confirm the reason with CEC.
- Not reported: a specification or document is missing or the source explicitly says no information was submitted. It does not mean zero, “no,” or failed certification.
Configurations stay together
We preserve each CEC model string verbatim apart from whitespace normalization. CEC annotations in braces identify voltage or capacity configurations; they are separated from the manufacturer’s model number for display. Those variants share a single canonical model page. Distinct manufacturer model numbers retain their own pages and evidence.
Ratings and source differences
Panel nameplate and PTC ratings are distinct DC watts. Inverter ratings are continuous AC kilowatts; storage capacity is kilowatt-hours. Legacy PCS apparent power is shown in kVA and is never converted to kW. A complex source rating remains text rather than an invented numeric value.
Repeated rows with identical identities are consolidated with every sheet and row reference retained. Distinct panel rating sets stay separate even if their model strings match. Hybrid inverter rows appearing on both worksheets share a record when ratings agree; conflicting ratings remain separate with an explicit source-discrepancy note.
Compatibility requires documentation
We do not match equipment based on electrical ratings, branding, or registry co-listing. A published compatibility claim must identify the counterpart, manufacturer document, evidence location, conditions, and check date. Evidence is scoped to the exact models and conditions shown. A model without verified pairing evidence is marked explicitly.
Coverage and source quality
The directory contains 35,051 tracked records across all seven supported active equipment lists. Counts include records retained after a later absence observation. Both solar and battery inverter worksheets are included. Service-provider lists (PMRS), the separate V2G program and historical CEC removal archives are outside this equipment catalog. Manufacturer documents supplement selected exact models; missing documents are marked on each page. CEC data includes manufacturer-reported information and is not an independent product test by SolarApprovals.
- Solar panels: 22,186 tracked records
- Inverters: 2,473 tracked records
- Energy storage: 6,837 tracked records
- Batteries: 991 tracked records
- Meters: 787 tracked records
- Power control systems: 94 tracked records
- Legacy PCS: 1,683 tracked records
The importer validates workbook layout, data types, row counts, and source checksums before a transaction. It upserts stable identities, preserves source snapshots, and adds a history event only when the observed status changes. Failed downloads and workbooks that fail validation cannot change observations. Minimum row counts and a 5% source-count decline guard help detect incomplete data, but do not prove upstream completeness.