Four coal plants. That’s what’s still keeping the lights on for eight million people in New South Wales. Bayswater, Eraring, Vales Point and Mount Piper between them supply somewhere near 60 per cent of the state’s electricity, and every one of those stations is pushing 30 to 45 years old. Liddell, the fifth, closed in April 2023. The other four were all supposed to be well into retirement by now, or close to it, according to the plan the NSW government legislated back in 2020. They’re not. And the reason they’re not tells you most of what you need to know about how the state’s transition is actually going, as opposed to how the press releases describe it.
The coal exit that keeps getting pushed back #
Eraring is the one everyone watches. It’s the biggest coal plant in the National Electricity Market, owned by Origin Energy, and it was originally slated to close in 2025. That date has already moved once, with Origin and the NSW government striking a deal to keep the plant running longer while the replacement firming capacity catches up. Origin’s own market filings describe the extension as a bridge, not a change of destination. Fair enough. But bridges have a habit of getting longer than planned.
Vales Point, owned by Sunset Power, doesn’t get the same headlines but its retirement horizon has quietly stretched too. Mount Piper, EnergyAustralia’s plant near Lithgow, is contracted to run into the 2040s under current plans, though EnergyAustralia has flagged that timeline is under constant review. AGL’s Bayswater remains the newest of the four and the least likely to close early. Follow the money here and it’s straightforward: none of these owners want to be the one holding a mothballed asset while the market is still short on dispatchable capacity. Our earlier look at coal closures and the reliability debate covers why AEMO keeps flagging the same risk nationally, not just in NSW.
EnergyCo and the REZ machine #
The legislative centrepiece of the state’s plan is the Electricity Infrastructure Investment Act 2020, and the delivery vehicle is EnergyCo NSW, a government agency built specifically to plan and build the transmission backbone for renewable energy zones. Four REZs are declared: Central-West Orana, New England, South West and Hunter-Central Coast. Central-West Orana, running through country around Dubbo and Wellington, is the furthest along and the most instructive case study, because it shows exactly where the friction sits.
The generation side of these zones moves fast. Wind and solar developers can get a project permitted, financed and built in a few years if the grid connection is there. The transmission side moves at a completely different pace, because building 500kV lines through farming country means easements, landholder negotiations, cultural heritage assessments and local council engagement, none of which compress just because a state government wants them to. EnergyCo’s own public updates on Central-West Orana have acknowledged timeline pressure and cost movement since the original scoping work. That’s not a scandal. It’s what infrastructure delivery looks like when the schedule was written by policy people rather than construction people.
HumeLink and the transmission bottleneck #
The single biggest chokepoint in the NSW plan isn’t a REZ line, it’s HumeLink, the Transgrid project meant to connect Snowy 2.0’s pumped hydro output into the grid via Wagga Wagga and Bannaba. Transgrid’s own project updates have flagged cost estimates rising well past the original figures and completion pushed out multiple times. Snowy 2.0 itself is worse. Snowy Hydro’s original public cost estimate was around $2 billion. The figure now sits past $12 billion by the company’s own disclosures, with completion expected later in the decade rather than the mid-2020s window first promised.
This matters for NSW specifically because Snowy 2.0 is meant to be the anchor firming asset behind the whole coal-exit strategy. Without it running at scale, and without HumeLink carrying the power out, the REZ rollout is building generation with nowhere reliable to send the firming capacity that makes it dispatchable overnight. We’ve covered the broader argument for and against pumped hydro against big batteries in pumped hydro vs big batteries: which firms the grid?, and NSW is the state where that argument has the most riding on it.
The Waratah Super Battery and the firming stopgap #
While Snowy 2.0 slips, NSW has leaned hard on batteries to buy time. The Waratah Super Battery, built on the old Munmorah power station site on the Central Coast, was designed as a system-security asset rather than a straightforward arbitrage play, there to stabilise the grid and hold Eraring’s exit open as an option without everything falling over in between. It’s one of the largest batteries in the country by the time it’s fully commissioned. It’s also a tell. When a government agency commissions a battery specifically to backstop the closure of the state’s biggest coal plant, that’s an admission the long-lead transmission and pumped hydro projects aren’t arriving on the original schedule. Readers wanting the mechanics of how big batteries actually get dispatched should look at our explainer on how the NEM dispatches power every five minutes.
Community pushback and the social licence problem #
Drive through the New England REZ corridor or the country around Merotherie and Uarbry near Central-West Orana and you’ll find landholders who aren’t hostile to renewables in principle but are furious about how easements, compensation and consultation have been handled. EnergyCo has revised its community engagement approach more than once in response. This isn’t unique to NSW, the same tension shows up around wind corridors in Victoria and Queensland, but NSW’s REZ model puts more transmission through more private agricultural land than most other states attempt at once, which raises the stakes on getting the social licence question right the first time.
The honest read is that NSW’s REZ framework is well designed on paper, arguably the most coherent state transmission policy in the country, and still running behind the timeline its own legislation implied. Comparing it to South Australia’s experience is instructive, and we’ve written about how SA’s high renewable penetration record hides its own trade-offs in South Australia’s renewable energy record: what it hides. NSW doesn’t have SA’s luxury of a small, simple grid. It has to move a much bigger, coal-heavy system without the price shocks that would come from moving too fast.
What the federal money is actually buying #
Layered on top of the state programme is the Commonwealth’s Capacity Investment Scheme, which is underwriting a chunk of the new generation and storage that’s meant to backfill coal’s exit. It’s worth being clear-eyed about what that scheme does and doesn’t guarantee, and our piece on whether the Capacity Investment Scheme is quietly picking winners goes into the mechanics. In NSW specifically, CIS-backed capacity is meant to arrive on a timeline that roughly tracks Eraring’s revised exit. Whether it actually does depends on the same transmission constraints described above, because a contracted megawatt sitting behind an unbuilt line isn’t delivering anything to Sydney.
AEMO’s Integrated System Plan keeps NSW’s REZ build classified as actionable and near-term critical, which in AEMO’s own language means the system genuinely needs it on schedule, not needs it eventually. That’s a useful check against the temptation to treat delays as background noise.
A state betting on timing it can’t fully control #
NSW’s plan is not a bad plan. It’s arguably the most structurally sound of any state’s, with a dedicated delivery agency, a legislated framework and four REZs doing real work bringing generation online. But plans meet physical constraints, and the physical constraints here are transmission builds, a pumped hydro project years behind its own schedule, and landholders who have every right to slow things down when they’re not consulted properly. My own reckon, having watched this rollout since the legislation passed, is that the 2030 targets attached to the roadmap will need a public rewrite within the next couple of years, the same way Eraring’s closure date already has. Nobody in Macquarie Street will say that out loud yet. The transmission workers in the Central-West Orana corridor already know it.
Worth remembering, too, that NSW’s household side of this equation, the rooftop solar and battery uptake covered in our piece on rooftop solar versus a home battery, is moving faster than the big infrastructure ever will. The grid-scale build is the hard part. It was always going to be the hard part. The question worth asking in twelve months isn’t whether Eraring’s closure date moves again, it’s by how much, and what backs it up when it does.
— Marcus Wren, Editor