Methods, assumptions and sources

Every number in this tool comes from a public source or a stated assumption on this page. Western Power publishes spare capacity in bands, and NCMT gives bands only; Western Power's TSP Appendix E gives exact utilisation for the most loaded (about 60) zone substations, which this tool uses where it exists. Elsewhere we say "needs ~X MW in an area with <5 MVA spare".

1. Matching a site to the grid

Each site is placed in the Western Power NCMT supply-area polygon that contains it (point-in-polygon). The polygon gives the capacity band at the site's location. A site's supply zone substation is named only where a planning or application document states it (cited on the lead); otherwise it is shown as unconfirmed. Distance to the nearest 132/330 kV line comes from Western Power's transmission line layer.

2. Time-to-power gap

gap = realistic grid-ready date − the date the customer needs power, always shown as a range (low–high).

3. MW confidence ladder and estimate rules

Badges: H stated in an official document · M stated by media or the proponent · L inferred, derived or typical-for-class estimate.

4. Sweet-spot test (this tool's screen), constraint rule and fit score

Sweet-spot test (this tool's screen) = connection under 10 MW on the load tested, load about 5 MW or less (definition; the tests applied are connection and grid draw) and grid draw about 4 MW or less, and the customer is a land developer (or estate land a developer will unlock). Judged on connection and grid draw, not total design load: a microgrid caps grid draw (Peel Business Park runs on a 2.2 MW microgrid today (Zenith Energy Peel project page: 'Capacity 2.2 MW')). Estates are tested on the next stage only; single sites on the full load. Where no connection or draw is published, connection = the load tested (the next-stage or first-stage design load for estates and staged sites; the full single-site load where no first-stage figure exists, labelled "Full load tested") and grid draw = 80% of the load tested (assumption, the ratio implied by 4 MW draw on a 5 MW load). (screening thresholds used by this tool)

Constraint rule. Constrained = Western Power NCMT remaining capacity of 20 MVA or less in 2029 (this tool's screening rule: augmentation likely to be needed); severe = under 5 MVA. Headroom is the filter, not the trigger: the trigger is a land purchase, rezoning, structure plan or release. (Western Power Network Capacity Mapping Tool)

The sweet-spot filter (this tool's screen; on by default) shows only leads that pass all three size tests, have a developer as the customer, sit in a constrained 2029 area and are still active. Oversized leads (data-centre campuses, the Alkimos desalination plant, estates whose first stage is 10 MW or more) move to the 'Too big / grid-stretch' tab. Standalone 10–100 MW loads score low unless they sit inside a developer estate; anything over 100 MW is capped at 50.

Sizing (design load, not metered demand). Industrial and commercial land: 0.20 MVA per net hectare of the next stage only (Western Power Underground Distribution Schemes Manual, 12th ed., cl. 5.3.3.3 (200 kVA/ha minimum for commercial and industrial lots, net lot area)), labelled "design load". Gross ULDO hectares are converted to net at 0.77 (assumption: Muchea, 896 net of 1,167 gross ha). Where no next-stage area is published the design load shows as pending and the size tests are skipped, not failed. Residential: 4.5–6.5 kVA per home, base case 5.5 kVA (Joint Supply and Planning Manual (STNW3470, Release 8, 7 May 2025), Table 24 (4.5 kVA); Endeavour Energy TB 0188A ADMD schedule (6.5 kVA)); both ends are shown. Western Power publishes no single per-lot figure: it sizes estates with its maximum demand (DADMD) calculator. Residential land given only in hectares: 0.10–0.14 MVA per hectare.

Microgrid logic. The fit score (this tool's screen) is judged on the connection and grid draw of the load tested, not total design load: a microgrid caps grid draw (Peel Business Park runs on a 2.2 MW microgrid today (Zenith Energy Peel project page: 'Capacity 2.2 MW')). Published connection or draw figures are used where they exist; otherwise connection = the load tested (the next-stage or first-stage design load for estates and staged sites; the full single-site load where no first-stage figure exists, labelled "Full load tested") and grid draw = 80% of the load tested. Every estimated grid draw is an assumption (screening assumption; actual draw depends on solar/BESS sizing; to be replaced by design studies) and is marked * in the lists.

Headroom edition. Where they differ, Western Power's TSP 2025 Appendix E (Jul 2026), PoE10 (2026–35, headroom = capacity × (1 − utilisation)) wins over older network data (e.g. Muchea about 6.6 MVA in 2029 rather than about 19). It covers 60 zone substations; elsewhere the Western Power NCMT capacity bands (2025 data) are used. Each lead card names the edition used.

Fit score (0–100)

Score = Σ weight × sub-score. Hard excludes: withdrawn, refused, cancelled, dead, idled or study-only projects, existing assets, unlocated sites.

Site signals (approved additions)

Score = Σ weight × sub-score + site-signal points, clamped to 0–100; transmission-scale leads are then capped at 50. Site signals in leads.json (rows with a fit score): rows with a fit score 242 · gas corridor within 1.5 km 60 · gas corridor 1.5–5 km 68 · wetland at the site 58 · contaminated-site record 31 · threatened ecological community (check with DBCA) 73 · not assessed (locality point only) 24.

Stage / timing factor ('call now')

The weighted score rewards fit; this factor rewards timing, so the top of the list is leads that can be acted on now. Points are added on top of the weighted score and site signals, then the total is clamped to 0–100 and rounded half-up (one rounded value, the same in the export and this app). Every flag has a source, shown on the lead card. Calibration sites are scored neutral (0) on this factor: they are scored as if planned today, so their own trigger dates are not comparable.

Sources of flags: published planning and application documents, the verification checks (team verification checks, 19 Oct 2026), and an automatic map match: a WAPC structure plan endorsed (new plans only; amendments to an existing plan earn no points) or an urban region-scheme amendment advertised inside a ULDO estate footprint in the last 12 months. ULDO records with no estate name are flagged 'no named developer'. Leads not checked are labelled inferred.

Source check

Leads are screened for early stage and single-controller land; each pin was checked by hand against public sources.

Hand-checked pins

Leads are screened for early stage and single-controller land; each pin was checked by hand against public sources. The two pins passed the verification checks (developer or landowner named by a primary source, dated trigger within 12 months, not already built, not already a licensed estate microgrid (ERA register)) and the source check. WATCH leads are badged, shown under the pins, and are not calls. Zone substations are labelled 'named in plan' only where a planning document names them (Stakehill, Wandi, Peelhurst/Sawley Close, and in the dropped tab North Ellenbrook West and Alkimos); so do the GreenSquareDC Hazelmere DAP agenda (no.141, p.403) and the South Bullsbrook LSP2 (2016, p.83). The approvals cards add Precinct 8 (Wanneroo), Port Kennedy (Waikiki, 2011 study) and Doobarda (Byford). CDC Maddington: no public planning document we opened confirms its supply substation (unconfirmed). Every other site's supply substation is shown as 'unconfirmed'; the NCMT polygon at its location gives only the capacity band.

North Ellenbrook West is now a weak pass, long-horizon watch, not a pin: the district plan (p.86) says the area is currently serviced by the Muchea Zone Substation, about 17 km north, and that a new zone substation is required at build-out; Muchea is not on the draft TSP 2026 highly loaded list. The district plan's "new zone substation required" line comes from a 2018 servicing report (Pritchard Francis, 26 Sep 2018; stale), and the first stage is tied to the Tonkin interchange (about 2028), so "first homes late 2027" is soft. The safe line: a future zone substation at 7,700–8,500 homes. It is shown as its stage 1 only: LSP Area 1, Gateway (South-East), about 120 ha (SWAN-SP/2026/2); the district plan names Muchea (TSP 2025 headroom about 10.2 / 8.7 / 6.6 MVA in 2027 / 2028 / 2029). The stage-1 yield is not public, so about 1,000–1,130 homes is an assumption (120 ha at the district average density), giving 4.5–7.3 MVA at 4.5–6.5 kVA per home. Stage 1 would exceed 10 MVA above about 1,818 homes at 5.5 kVA. At the base case its estimated grid draw (assumption) is just over 4 MW, so it tests as 'edge' rather than core. Trigger: the WAPC resolution of 16 Sep 2026 to lift urban deferment on 325 ha (aggregator source; the Government Gazette notice was not re-opened). Amendment 2 proposes 7,700–8,500 homes for the whole district.

Calibration against licensed estate microgrids (ERA register)

The estate and park microgrids named on distribution licences EDL7, EDL9 and EDL10 in the ERA electricity licence register are scored with exactly the same rules as live leads, as if planned today. They are a calibration check only, not leads, and are not shown in the lead lists or on the map. Industrial estates are sized on the next stage only at 0.20 MVA per net ha (Busselton Airport: the 15 ha precinct being marketed, not re-verified; the ERA licence area there is 165 ha; Busselton Home HQ: 15 ha from EV-Energy's ERA application, whose summary table says 24 ha; Muchea: EV-Energy's ERA application publishes Stage 3 (8 lots pre-sold, first customers Sep 2026) and Stage 4 (18 lots, 2026) but no stage areas, so the about 80 ha still unsold is used as an upper-bound ESTIMATE, giving about 16 MVA design load, over the 10 MW connection limit, so Muchea tests 'over' and sits outside the sweet spot; it fits only if a stage of 25 net ha or less is published (25 × 0.20 = 5 MVA); Peel: Stage 3 hectares are not published, so its design load is pending and it is judged on the 2.2 MW microgrid capacity published by Zenith Energy (project page); DevelopmentWA names Peel Renewable Energy as the long-term operator). Zenith-group rows are disclosed only where a source shows it: Peel (a Zenith Energy company) and the EV-Energy rows (a Zenith Energy subsidiary, per its ERA application); Ocean Reef ownership is not attributed (no source opened shows it). Residential precincts use 5.5 kVA per home (range 4.5–6.5): Ocean Reef Marina on first-stage homes with the published 4 MVA connection; Eglinton Village on the district-centre structure plan dwellings (an assumption to confirm). Rank is the position the site would take among the 173 scored rows (draft: 167): 1 + the number of those rows scoring strictly higher, so sites with nearby scores can share a rank. Calibration sites are neutral on the stage / timing factor (their own triggers are years old), while live leads gain or lose up to +10 / −25 points for timing. Muchea being outside the sweet spot is expected and kept as it is.

Rank = the position each licensed site's score would take among the 174 scored rows it is ranked against: the 173 target and estate rows on TSP 2025 that have a fit score and are not in Too big, Not a microgrid lead or Unresolved, plus the site itself. Industrial estates are sized on the next stage only and judged on connection and grid draw (* = grid draw is an assumption).

Each site is ranked separately against the pool, so sites with close scores can share a rank.

Alert sources

5. Value context

Not shown in this build: no monetary value figures are displayed.

6. Caveats

• NCMT bands are load headroom; for generation and storage they are a proxy only.
• Headroom is a planning forecast (PoE10), not a connection offer.
• Whether NCMT already includes Western Power's planned changes is not verified in this build; NOM projects are shown as a cross-check only, and a NOM 'Required by Date (subject to change)' is a need date, not a delivery date.
• Western Power gives connection-ready projects designated under its Critical Projects Framework (launched May 2024) 'right of way to Western Power resources'. Our assumption: projects not designated critical may wait longer than the average connection time, so our grid-ready dates may be optimistic. (Western Power, Critical Projects Framework page (checked 21 Oct 2026))
• Geocodes at locality level are coarse and cap the evidence score.
• Licences: the SLIP map services are licensed for personal, non-commercial use unless a dataset's own licence or a SLIP subscription covers it, and the DBCA wetlands (DBCA-019) and TEC (DBCA-038) layers are CC BY-NC 4.0. That is fine for this demo, but a paid product needs these licences cleared. The petroleum pipelines layer (DMIRS-007) is CC BY 4.0.
• EPA referrals do not carry MW as data: the EPA referred-proposals layer (DWER-120) has no MW, generator or emissions field. MW sometimes appears in the referral PDFs (e.g. greenhouse gas plans), so we use referrals only as a lead-finding signal, read by hand, never as an automatic MW source.
• Gas corridors are licence corridors, not supply: a tap off DBNGP or Parmelia needs the pipeline owner's agreement (DBP or APA), a gate station and a lateral, plus supply contracts, so allow about a year or more. Smaller sites would more likely use the ATCO distribution network, which is not mapped here.

7. Data freshness

Exported 2026-10-19T22:41 AWST · refresh status ok (20261019_222234). Live WEM prices, when available, come from the host at ./api/oe/market.