Most constrained wind farms in the NEM, ranked

A live dashboard of Australia's National Electricity Market, built with NEM Explorer and refreshed from AEMO data on every view · updated 31 October 2026

Most constrained wind farms in the NEM

Every wind farm of 20 MW or more, dispatched as a semi-scheduled unit, in Australia's National Electricity Market (NEM), ranked by the share of its available energy that it spilled. Available energy is what the farm could have generated in the period with nothing holding it back. Spilled energy is the part of that it was not dispatched to generate. Each bar is one farm, and the bars fall from left to right.

Tick Network curtailment, Economic offloading, or both, to choose which losses count as spill. Network curtailment is output that grid limits held back; economic offloading is output the farm held back itself because the spot price was below its offer. Pick the period with the time range at the top of the page, a region with Region, and how many farms to show with Farms shown. The figures cover whole days up to yesterday.

Ranked wind farms, selected period

Figures as of 8 November 2026, 1:43 am NEM time.

• 83 wind farms ranked
• 695.3 GWh lost to network curtailment
• 2,891 GWh lost to economic offloading
• 9.897% of their available energy spilled
• 19.39% of all lost energy was network curtailment
• 6.505% median farm's spill, % of its available energy

Share of available energy spilled, per farm

Spill counts the losses ticked above. Tick at least one; the table splits each farm's spill into network curtailment and economic offloading.

Most constrained wind farms: the ranking

Spilled counts the ticked losses; the network and economic columns always show both. The GWh columns give each farm's energy before dividing, so a small farm with a high share stands out from a large one.

Network curtailment and economic offloading

These are AEMO's own definitions, the ones its Quarterly Energy Dynamics reports use.

• Network curtailment is output the farm offered below the regional spot price that dispatch still did not take. A network or system security limit held the farm back, so it lost energy it wanted to sell at that price.
• Economic offloading is output the farm offered at or above the spot price. It was not dispatched because the spot price was below the price the farm asked. The farm chose this through its offer.
• Spilled is whichever of the two the tick boxes select. With both ticked it is all the energy the farm could have generated but was not dispatched.
• Ranking by share compares farms of different sizes, and in different places, fairly. A large farm loses more energy in total; dividing by the energy it had available shows how hard each one is hit. The table gives the losses in GWh as well.

For the same losses across every wind and solar farm, by region and month, see NEM wind and solar curtailment by region. For the same ranking of solar farms, see Most constrained solar farms.

Build a dashboard like this

An AI agent built this page from a plain-English request using NEM Explorer, which holds AEMO market data from 2009 onward. The page re-runs its queries against the latest data each time it is opened. Ask for your own at chat.scadaminer.com, or connect Claude Code or another MCP client to NEM Explorer and have it save one for you. The request behind this page was close to:

Build me a dashboard ranking NEM wind farms by the share of their available energy spilled, with one bar per farm ordered high to low, tick boxes to count network curtailment and/or economic offloading, and a date range picker.

Method, sources and caveats

• Available energy is AEMO's dispatch availability for each farm in each 5-minute interval, added up over the period. For a semi-scheduled wind farm it is the lower of the availability the farm offered and a forecast of what the wind allows. That forecast is AEMO's, or the farm's own where AEMO uses it. Generated is its dispatch target, the output it was told to produce; metered output differs slightly. Lost is availability minus the dispatch target, floored at zero in each interval.
• The split follows AEMO's method. The farm's energy offer for the interval is capped at its availability, filling the lowest-priced bands first. Each band's price is divided by the farm's marginal loss factor so it compares with the regional price. The capped MW offered below the regional price is the farm's economic availability. Network curtailment is economic availability minus the dispatch target, when that is positive. Economic offloading is the rest of the loss.
• Shares are energy as a percentage of the farm's available energy over the period. A wind farm's available energy rises and falls with the wind at any hour, so calm periods count on neither side.
• A farm that lowers the availability it offers, instead of offering its output at a high price, removes that energy from both sides of the share, so its economic offloading reads low.
• Farms included: semi-scheduled wind units of 20 MW or more registered capacity, with dispatch data on every day of the period, that had already reached 70% of their usual peak availability early in the period. That leaves out farms still commissioning or that opened or closed during it, so a range of several years ranks fewer farms. The range can reach back to July 2009, where this data starts, and spans at most eight years. A farm registered as several units shows one bar per unit. Non-scheduled wind farms, such as Waubra, Capital and Woolnorth, are not dispatched and have no availability or offers in AEMO's data, so they are not ranked.
• Offers for trading days from March 2021 to July 2024 are rebuilt from AEMO's offer-tracking tables, because AEMO did not publish the offers in effect for those days.
• Data is from AEMO's dispatch unit solution, dispatch price and bid tables, in NEM time (UTC+10).

Contains material sourced from AEMO. SCADA Miner transforms and analyses the source material; results are not endorsed by AEMO.

Ask your own question of NEM data

More analyses

More published analyses and dashboards