NEM Wind & Solar Curtailment — measured

AEMO's own per-farm availability forecast (UIGF) minus what each farm actually generated · Dec 2025 – Aug 2026

Overview
Time of day
Regions
Farms
Compare
▸ How this is measured — and what it is not

For every semi-scheduled wind and solar unit, AEMO publishes an unconstrained intermittent generation forecast (UIGF) for each 5-minute dispatch interval — a weather-driven estimate of what the farm could have produced. Subtracting metered SCADA output gives energy that was available and not generated.

This is gross curtailment. UIGF is not capped by the farm's bid, so the figure covers energy withdrawn for commercial reasons as well as energy the network refused. It is not comparable with network-only curtailment figures.

Days a farm did not run at all are excluded — that is plant unavailability, not curtailment. Days a farm ran at a stable fraction of its forecast are flagged as likely derates and reported separately, since a proportional shortfall suggests part of the plant was out. A per-unit threshold of max(1 MW, 2% of peak forecast) suppresses ordinary forecast error; because intervals where a farm outgenerated its forecast are not netted off, the remaining figure is an upper bound.

Yesterday — 2026-09-11

Computed each build from AEMO's previous-day per-unit forecast against metered output. 188 farms.

Curtailed
11,777 MWh
8.8% of available
Available
134,084 MWh
forecast, unconstrained
Most curtailed
WLWLSF1
846 MWh
Period curtailment
11.6%
Dec 2025 – Aug 2026
Energy curtailed
4,907 GWh
of 42,322 GWh available
Worst month
20.7%
Dec 2025 · complete months only
Farms measured
187
constant fleet across months

Curtailment by month

Share of forecast-available energy not generated. Dashed line = the same figure with likely-derate days removed; the gap is how much rests on that classification.

1 incomplete month on the chart. Months with under 90% interval coverage are hatched and excluded from the figures above — a part-month ratio has a denominator a fraction the size of a full one, so it swings wildly and is not comparable. If an incomplete month is not simply the current one still filling, the data for it did not download.

Wind vs solar

Curtailed energy by technology, and each technology’s own curtailment rate. Solar dominates the volume in summer; wind carries the overnight and shoulder hours. The two labels are each measured against that technology’s own available energy, so they do not add to the fleet rate above — that figure is the energy-weighted average of the two, and sits between them. Bar heights (GWh) do add, up to any farms whose technology is unmapped. Incomplete months are hatched here too.

Read this as gross curtailment. It includes farms choosing not to generate through negative prices as well as network constraints, and the split between those two is inferred from the observed regional price, not from which constraint bound.

When curtailment happens

MWh curtailed per half-hour slot. The midday peak is the solar signature — every farm generates at once, prices fall, and network limits bind.

Curtailment by region

Share of each region's forecast-available wind and solar energy that was not generated, over the whole period.

Compare assets

Pick up to four farms. Filter by name, technology or region, then click to add or remove. Comparison is over the whole period shown in the header.

Side by side

Totals over the period. Curtailed % is the share of each farm's own forecast-available energy.

Curtailment rate by month

Each farm's own rate, so a small farm and a large one are directly comparable.

When each farm is curtailed

Share of that farm's total curtailment falling in each half-hour. Normalised, so the shapes compare regardless of farm size.

By farm

Top 60 by energy curtailed. Click a column to sort. Tags mark units whose behaviour was classified: an export cap, a farm cut to zero rather than trimmed, or a suspected partial outage.

DUIDTechRegion Available MWh Curtailed MWhCurtailed %