Four coal units. That’s what NSW has taken out of the grid since Liddell shut in 2023, with Eraring’s four units now running on a negotiated extension that was never meant to be permanent. Origin Energy’s own filings describe the Eraring arrangement with the NSW government as a bridge, not a destination, and bridges have a habit of being shorter than the traffic wants them to be. The state’s coal exit is the fastest of any NEM region relative to its remaining fleet. The question nobody in Macquarie Street has answered cleanly is whether the Renewable Energy Zones built to replace that capacity are moving at anything like the same pace.
They are not. Not yet, anyway.
What’s actually left of the coal fleet #
NSW still runs four coal stations of consequence: Eraring, Bayswater, Mt Piper and Vales Point. Between them they supply the bulk of the state’s firm capacity on any given evening, and AEMO’s own closure-year assumptions have all four winding down across the back half of this decade and into the early 2030s, Eraring aside, which now has government backing to keep going past its original 2025 retirement date. I’ve written before about the mechanics of that Eraring deal and the underwriting arrangement behind it – see Eraring at 2027: the deal, and the short version is that the extension buys time, not certainty. EnergyAustralia’s Mt Piper and Tallawarra sit in a similar holding pattern, covered in EnergyAustralia: the gentailer counting down to Yallourn’s exit, though that piece is really about the Victorian asset. The point stands across both gentailers: the coal fleet is being kept alive through policy intervention rather than retired on schedule, because the replacement generation isn’t ready to carry the load.
That’s not controversial. AEMO said as much in its most recent Integrated System Plan update, and the state government’s own energy roadmap documents have quietly pushed assumptions around at the margins for two years running.
The REZ rollout, honestly assessed #
NSW has five declared Renewable Energy Zones on paper. Central-West Orana is the furthest along and still the best test case for how this is actually going. I covered the detail in Central-West Orana REZ: NSW’s first REZ, and even there, transmission easements, landholder negotiations and connection queue bottlenecks have stretched timelines well past the original schedule. New England REZ, built around the transmission backbone that AcenRenewables and others are developing, tells a similar story; see New England REZ: where the wind-solar-storage build actually stands for the specifics on where wind, solar and storage commitments actually sit versus the slide decks from a few years back.
The pattern across both zones is consistent. Generation capacity, the wind and solar projects themselves, can be built in two to three years once financed. Transmission takes five to eight, sometimes longer once you account for planning approvals, landholder compensation disputes and the physical task of stringing steel across grazing country. NSW’s REZ transmission build is running on the longer end of that range almost everywhere. EnergyCo NSW, the state body coordinating the REZ rollout, has acknowledged delays publicly in its project updates, citing supply chain constraints for transformers and high-voltage cable alongside the usual approvals friction.
Why the gap matters more in NSW than elsewhere #
Every NEM state is juggling some version of this mismatch. South Australia solved most of its version of the problem a decade ago through a mix of wind, batteries and interconnection, and I’ve gone through why that state keeps setting precedents the rest of the NEM eventually follows in South Australia: the grid that keeps teaching the NEM lessons. Queensland is running its own version with pumped hydro as the backstop, discussed in Queensland’s energy plan rests on two pillars and one cliff face.
NSW’s problem is scale. It’s the largest load centre in the NEM, Sydney alone accounts for a demand block that dwarfs most other capital cities, and the coal stations doing the heavy lifting today are old enough that unplanned outages are becoming a routine feature of the operating environment rather than an exception. AEMO’s reliability forecasts have flagged NSW as carrying tighter reserve margins than most other regions through the middle years of this decade. When Bayswater or Vales Point trips unexpectedly on a 40-degree February afternoon, the REZ transmission that’s supposed to be moving New England wind and Central-West Orana solar into Sydney isn’t finished enough to fully absorb the shock.
The firming question underneath it all #
Wind and solar generation without the wires to move it is just stranded capacity. This is where the REZ delay compounds rather than simply adds to the problem. Developers who signed up for transmission access in the connection queue are sitting on finished or near-finished projects waiting for substations and lines that aren’t there yet. The Waratah Super Battery, built specifically to act as a shock absorber for the wider grid rather than a conventional REZ asset, has started proving its worth on exactly this kind of volatility – see Waratah Super Battery: the shock absorber finally earning its keep, but one big battery, however well engineered, isn’t a substitute for finished transmission corridors.
Gas peaking capacity fills some of the gap in the meantime, and Kurri Kurri’s long and awkward path to commissioning is instructive here, covered in Kurri Kurri: the peaker that took the long way round. The broader argument about what “firming” even means in a system this stretched is worth revisiting too, in What ‘firming’ means and why gas keeps coming up. My own read is that gas gets blamed for being a crutch when it’s actually functioning as designed: a short-duration backstop for exactly the kind of gap NSW has created by letting transmission lag generation. The criticism should be aimed at the timeline mismatch, not the technology filling it.
Who actually pays for the delay #
Transmission cost recovery in NSW runs through the regulated revenue framework the AER oversees, and I’ve gone through the mechanics of who ultimately carries that cost in Who really pays for the transmission build-out. The short practical answer for NSW specifically: consumers absorb the REZ build costs through network charges over decades, but they also absorb the cost of coal running longer than planned, because an ageing fleet needs more maintenance capital and because scarcity pricing during tight periods flows straight through to wholesale costs. Both sides of this ledger land on the same power bill eventually. Nobody at EnergyCo or AEMO is pretending otherwise in public forums, even if the political messaging sometimes suggests the REZ build is free money sitting on the table.
It isn’t. Nothing about rebuilding a state’s transmission backbone while running four ageing coal units past their use-by date is free.
What a realistic timeline looks like #
Push past the official project milestones and talk to people working the connection queue, and the consensus settles on the early-to-mid 2030s before NSW’s REZ transmission genuinely matches the capacity it’s meant to carry. That’s roughly in line with when Bayswater and Mt Piper are pencilled in to exit under current closure assumptions, which sounds like it lines up neatly. It mostly does, on paper. The risk sits in the gap years immediately before each closure, when reserve margins are thinnest and any unplanned outage at a remaining coal unit collides with a REZ that’s still three-quarters built. AEMO’s own Electricity Statement of Opportunities has flagged NSW reliability gaps appearing earlier than the broader ISP timeline assumed, which is the kind of warning regulators tend to publish quietly and regret loudly later.
I’ll admit the state government’s messaging on this has gotten more candid over the past year, which counts for something. EnergyCo’s public project trackers are more honest about delays than they were two years ago, and that’s a better starting point than pretending the schedule is intact. Still, honesty about delay isn’t the same as closing it.
The bit that gets underweighted in the public debate #
Most of the political argument about NSW energy policy treats the coal exit and the REZ build as if they’re the same project running on the same clock. They’re not. One is a retirement schedule largely determined by asset condition and owner economics. The other is a construction programme at the mercy of planning approval timeframes, landholder negotiation and a global supply chain for transformers that every state and plenty of overseas grids are competing for simultaneously. Treating them as synchronised was always optimistic. NSW’s energy planners know this. Whether the public conversation catches up before the next unplanned coal outage makes the point for them is the open question worth watching into 2027.
For readers wanting the state-by-state comparison on how NSW’s REZ ambition stacks against other jurisdictions’ renewable targets, the scorecard is worth a look: State renewable targets compared: the honest scorecard. The AEMO Electricity Statement of Opportunities (aemo.com.au) and the NSW government’s own Electricity Infrastructure Roadmap publications, available through the state’s energy department (energy.nsw.gov.au), are the primary sources behind most of the timeline detail above, and both are worth reading in full rather than taking a journalist’s word for it.
– Marcus Wren, Editor
Photo by Artem Kniaz on Unsplash