Three clocks

It is easy to talk about network data as though availability were one property. It is at least three: how often a record refreshes, how finely it resolves, and what you must agree to before you can read it. Western Power's published record separates cleanly on all three, and the separations do not line up.

The interactive Network Opportunity Map is the fast layer. Western Power states that "the data presented in the iNOM is updated daily (8:30am AWST), with spatial data available for download", and describes it as a map-based view of capacity and emerging opportunities at zone substations and feeders across the SWIS. Daily cadence at feeder level is the finest-grained public reading of this network that exists.

The Network Opportunity Map itself and the Transmission System Plan are the slow layer, produced annually. Western Power publishes the NOM under the Electricity Network Access Code and the TSP under the Electricity System and Market Rules, so the annual cadence is a regulatory artefact rather than an editorial choice. The NOM gives a five to ten year performance snapshot, and its stated primary purpose is "to provide opportunities available to businesses to become involved in providing non-network solutions to the grid, such as demand management, battery storage options or an alternative option".

The third layer is the catalogue. A CKAN package_search on the western-power organisation at data.wa.gov.au returned 36 datasets on 14 August 2026. The asset layers in it carry a monthly update frequency, and the five sampled on that date all reported a last-modified timestamp of 14 August 2026.

Compiled from Western Power's Transmission System Plan and Network Opportunity Map page, its Network Capacity Mapping Tool page, and dataset records at data.wa.gov.au, read 14 August 2026. Cadence values are as stated by the publisher, not observed refresh behaviour.

The forward series

The part of this record I would open first is the one that took the longest to find. The Network Capacity Mapping Tool publishes forecast remaining capacity as a series of separate annual datasets, each showing, in Western Power's description, the "forecast remaining capacity of Western Power substations for the specified forecast year" as coloured polygons over supply catchment areas, in MVA segments, and including planned future substations.

They are individually catalogued and individually coded. A catalogue search on the title "Forecast Remaining Capacity" returned 13 datasets on 14 August 2026. Six were opened directly, and their codes advance with the year.

Dataset records at data.wa.gov.au, read 14 August 2026. The code series advances one step per forecast year, which is consistent with an unbroken annual run between the years listed, but the intervening editions were not opened and are not asserted here.

This is the layer that sits in the gap. AEMO publishes the SWIS at system level. The Clean Energy Regulator publishes installation and capacity records by postcode. Neither resolves to the substation catchment, and neither looks forward by edition. A forward capacity reading per catchment, republished each year, is a different kind of record from both, and it is the one that answers where on this network there is room.

The layer that went dark

One dataset in the catalogue is marked retired. WP-050, "NCMT Substation Generation Connection Capacity", was described as displaying indicative information to visualise the ability of the SWIS, as it currently exists, to connect extra mid-sized generation, including renewable energy projects. Its last update is recorded as 13 October 2022, its frequency as yearly, and the catalogue entry directs users to contact Western Power for additional information. No API is offered for it.

The timing is worth stating plainly, and then stating carefully. The Commonwealth announced on 5 August 2026 that the Small-scale Renewable Energy Scheme threshold rises from 100 kW to 1 MW, with commencement expected on 1 October 2026 subject to the necessary regulations being made. Solar Analytica already records that change as announced rather than in force. If it commences, the size class whose economics move is mid-sized generation, and the public layer that described this network's capacity to connect mid-sized generation stopped being updated nearly four years ago.

I am not claiming those two facts are connected. A retirement in 2022 cannot be a response to an announcement in 2026, and the layer may have been folded into the current mapping tool rather than withdrawn. What I am claiming is narrower and checkable: at the moment a size class becomes commercially interesting, the most directly relevant published layer for it is four years stale, and nothing in the catalogue entry says what replaced it.

Public, and gated

The access terms are the property most likely to be misread, because "on the open data portal" reads as open. The Western Power datasets carry a custom licence: Western Power's own data licence terms and conditions, with registration and agreement required before access. Services do exist for the live layers, including Web Feature Service, Web Mapping Service and ArcGIS map and feature services, but the catalogue records that they require login credentials, with token authentication on the ArcGIS feature service.

That is a materially different posture from the sources this publication leans on most. Clean Energy Regulator postcode files and AEMO's market data are retrieved without an account. Western Power's are not. For anyone building a data product on this network, the binding constraint is not cadence and not granularity, both of which are better than expected. It is the licence, and the work it implies is administrative rather than technical.

What we read, and what we do not

Applying this to our own record is the uncomfortable part. Solar Analytica's SWIS DER snapshot lists six sources. Exactly one is from Western Power: the Transmission System Plan and Network Opportunity Map page, the annual document. It supports the network opportunity figures the snapshot carries, 15 opportunities totalling 102.06 MW of active power, and the snapshot already warns that those mapped polygons do not imply uniform export capacity or connection constraint inside them.

So our WA network layer refreshes annually because we chose the annual document, while the operator republishes the finer version of it every morning at 8:30. The daily iNOM, the forecast capacity editions and the 36 catalogued datasets are all currently unread by anything we publish. That is a sourcing decision we made by default rather than on the merits, and it is the finding I would act on first.

What I am not willing to conclude

Four limits, and the first governs everything above. I have not accessed any licensed Western Power dataset. Every statement here rests on catalogue metadata and Western Power's public descriptions, so field definitions, actual spatial coverage, update reliability and data quality are all unverified. A catalogue that says monthly is not an observation that it refreshed monthly.

Second, the two counts are live values with a timestamp on them, not stable facts. Thirty-six datasets under the western-power organisation and 13 titled "Forecast Remaining Capacity" were the values returned on 14 August 2026, retired entries included in both.

Third, the forecast series is verified at six years, not across its full run. The code series is consistent with annual continuity, but consistency is not confirmation.

Fourth, I cannot tell from the catalogue whether WP-050 was withdrawn or superseded. That is a question for Western Power rather than for the portal, and it is the one I would ask first, because the answer determines whether there is a real gap in the public record for mid-sized generation connection capacity or merely a moved file.