The baseline comes first
The Clean Energy Regulator source edition records 564,139 allocated small-scale solar installation records representing 3,167.33 MW across the public SWIS extent through May 2026. Its newer battery register records 43,150 installation records and 933.64 MWh of usable capacity. Those are certificate-based records, not unique connected premises: replacements, upgrades and off-grid systems can be included, while certificates may be created up to twelve months after installation.
3.17 GW + 0.93 GWh
Allocated small-scale solar capacity and usable battery energy in the May 2026 CER source edition.
The observation boundary matters because the most recent twelve-month solar cohort contains 31,021 records against 33,498 in the preceding twelve months. That difference is not treated as a settled market contraction. The newer period remains incomplete under the source's validation mechanics.
A 2030 range, not a forecast dressed as one
The solar envelope compounds the May 2026 baseline through December 2030 using three annual net-addition corridors: approximately 160 MW in the lower case, 275 MW in the reference case and 360 MW in the upper case. The result is 3.9-4.8 GW, with 4.4 GW as the reference planning case. These rates are transparent sensitivities selected to test infrastructure, orchestration and market-service requirements; they are not assigned probabilities.
| Metric | May 2026 observed | 2030 lower | 2030 reference | 2030 upper |
|---|---|---|---|---|
| Allocated small-scale solar capacity | 3.17 GW | 3.9 GW | 4.4 GW | 4.8 GW |
| Battery installation records | 43,150 | 120,000 | 250,000 | 420,000 |
| Allocated usable battery energy | 0.93 GWh | 3.2 GWh | 5.5 GWh | 9.2 GWh |
| Battery records per 100 May-2026 solar records | 7.6 | 21 | 44 | 74 |
The final row uses the frozen May 2026 solar-record base as a common scale. It is not a forecast of household attachment rate.
Storage is the wider uncertainty
The battery source history begins only in July 2025. The 2030 endpoints therefore function as adoption and orchestration stress cases rather than an extrapolated statistical forecast. The lower case reaches 120,000 records and 3.2 GWh; the reference case 250,000 and 5.5 GWh; the upper case 420,000 and 9.2 GWh. Those combinations imply approximately 26.7, 22.0 and 21.9 kWh per record respectively. That arithmetic is a consistency check, not evidence of a likely system size, and no case asserts what share will be coordinated, dispatchable or available at a system peak.
The distinction is operationally important. Installed energy is not equivalent to peak-demand reduction, a market service or reserve capacity. The Forecast Accountability record preserves AEMO's separately published coordinated-battery assumptions. This scenario layer does not substitute its own endpoints into AEMO's capacity model.
What would reopen the range
Four triggers are declared in advance: a complete new CER source year that moves the mature annual solar run-rate outside the corridor; twelve additional months of comparable battery records; a material change in published VPP participation or availability; or a network-access reform that changes export and orchestration constraints. A crossed trigger opens a review. It does not rewrite this July 2026 edition.
The practical use is capacity planning. A network, retailer, aggregator or local-government team can test whether its 2030 program remains useful at 3.2 GWh or 9.2 GWh of distributed storage, then retain the same source record when the next edition arrives. The live WA / SWIS market profile carries the latest evidence; the fixed market edition and this Projective Layer preserve what was known and assumed at publication.
Boundaries
- The geography is the public SWIS allocation used by the Solar Analytica atlas, not every Western Australian electricity connection.
- CER certificate records are not unique premises and recent periods remain incomplete.
- Battery history begins in July 2025, limiting trend inference.
- The envelope does not model tariff changes, equipment prices, household income, network hosting capacity or policy intervention independently.
- No scenario endpoint is an AEMO forecast, a confidence interval or a most-likely outcome.