WA’s SWIS: the grid with no neighbour to call

Drive two hours south of Perth down the Forrest Highway and you’ll pass transmission towers carrying power that, physically, cannot go anywhere else in the country. No interconnector runs east to the NEM. No emergency import from Victoria or South Australia when a big unit trips. The South West Interconnected System serves roughly a hundred thousand square kilometres and about two and a half million people, and whatever generation, storage and demand response exists inside that footprint is what WA has got, full stop.

I’ve written before about the mechanics of the isolated SWIS grid and why that isolation matters more than most people appreciate from the east coast. Today I want to walk through where the system actually sits heading into summer 2026-27, because the settings have shifted since I last checked the AEMO Wholesale Electricity Market data for the region.

Why isolation changes the maths #

In the NEM, a sudden plant trip gets absorbed partly by flows across Heywood or Murraylink or the Queensland interconnectors. In the SWIS, there’s no equivalent release valve. Every megawatt of reserve has to exist inside the fence. That’s why AEMO, which also runs the Wholesale Electricity Market for WA under a separate set of market rules to the NEM, treats reserve margins with a different kind of caution. The relevant planning document is the Electricity Statement of Opportunities for the WEM, and it’s worth reading alongside the NEM’s Integrated System Plan if you want to see how differently the two systems are being steered.

Let’s be careful with the comparison, because the SWIS is a genuinely smaller system – peak demand sits somewhere around four gigawatts, a fraction of the NEM’s scale. But smaller doesn’t mean simpler. A forced outage on one large unit represents a much bigger proportional hit to supply than the same outage would in NSW or Queensland. The system has to carry more headroom per capita, and that headroom costs money that WA ratepayers and the state government, through Synergy and Western Power, ultimately wear.

Coal’s retreat on a different timetable #

WA is closing coal too, just on its own schedule and largely out of the national conversation. Synergy’s Muja and Collie power stations have been stepping down units for years, and the state government’s position, set out in its own energy transformation strategy, is for coal-fired generation to exit the SWIS by 2030. That’s a tighter deadline in calendar terms than parts of the NEM, even though the public debate about coal closures tends to focus on NSW and Victoria.

The difference is what replaces it. On the east coast, closing coal plant leans on transmission-connected renewable energy zones feeding a market that can also lean on interconnection in a pinch. In WA, the replacement has to be self-sufficient from day one. There’s no fallback import if the new solar, wind and storage don’t turn up on time. That’s a harder commitment than it looks on a press release, and I’d argue it deserves more scrutiny from analysts who treat the SWIS as a rounding error next to the NEM’s gigawatt numbers.

Rooftop solar’s outsized role #

Perth’s rooftop solar penetration is among the highest of any major Australian city, and in the SWIS that shows up directly in minimum demand troughs that have become genuinely difficult to manage. On a mild, sunny weekend day, underlying demand can fall low enough that AEMO has had to curtail rooftop solar output remotely under its emergency solar management scheme, something the NEM doesn’t do at anything like the same scale. I’ve touched on the broader version of this problem before: the question of whether rooftop solar is quietly breaking the grid applies with extra force in an island system that can’t export the surplus anywhere.

Western Power’s network, the poles-and-wires business separate from Synergy’s generation and retail arm, has had to invest heavily in dynamic export limits and smarter inverter settings just to keep voltage within bounds on low-demand days. It’s the distribution-level version of a problem the whole NEM is wrestling with, but WA hit it earlier and harder because there’s no bigger grid next door to soak up the excess.

Batteries doing more than firming #

The state government’s big storage bet, built mostly through Synergy-contracted batteries at sites like Kwinana and Collie, isn’t just playing the arbitrage and firming role batteries play in the NEM. In an island system, grid-scale batteries also have to help with frequency control and system strength functions that would otherwise fall to synchronous coal and gas units. As those coal units retire, the batteries and whatever grid-forming inverter capability comes with them become load-bearing for system stability, not just for shifting energy around the clock.

Worth being precise here, because the batteries themselves report power and energy capacity separately: a 200 MW battery with four hours of storage behind it is a very different asset to a 200 MW battery with one hour, and the SWIS press releases don’t always make that distinction as clearly as AEMO’s own data does. I’d point readers back to the general discussion of gas peakers versus big batteries in the firming job for the underlying logic; the same trade-offs apply in WA, just without a gas-fired interconnector two states away to lean on if the batteries run short.

Gas still carries a lot of the load #

It’s easy to forget, reading east-coast coverage, that gas is a much bigger piece of the WA generation mix than it is in most of the NEM states, partly because WA sits on its own large domestic gas reserves and has a reservation policy requiring a share of production to stay onshore. That’s a genuinely different starting position to the east coast’s gas supply squeeze I’ve covered in the context of winter 2028. Gas peaking and mid-merit plant in the SWIS isn’t a stopgap waiting to be phased out in the same urgent tone you hear about Kurri Kurri or Tallawarra B; it’s treated more as a structural partner to the renewable build, at least for now.

Whether that holds as batteries scale up is the open question. My honest read is that WA’s gas fleet has a longer runway than equivalent plant on the east coast, mostly because the state doesn’t face the same export-driven domestic gas pressure that Queensland and the southern states do. That’s a contrarian thing to say when the national mood is “gas is finishing up everywhere,” but the WA numbers don’t support that timeline, at least not yet.

What the transition actually needs to prove #

The SWIS doesn’t have a Marinus Link or a VNI West waiting in the wings to rescue a bad planning year. Every renewable energy zone, every battery, every bit of transmission reinforcement inside WA has to be sized for a system that can’t ask a neighbour for help. That’s a genuinely different engineering problem to the one NSW or Victoria are solving, and I think it gets underweighted in the national conversation simply because the SWIS is smaller in absolute terms.

I keep thinking about this the same way I think about a long solo ride with no support vehicle. You carry your own spares, you plan your own fuel stops, and there’s no truck coming to pick you up if something fails forty kilometres from anywhere. WA’s grid planners are making exactly that kind of provision, just with megawatts instead of water bottles. The AEMO Electricity Statement of Opportunities for the WEM and the state government’s own energy transformation documents are the places to watch for whether that provisioning keeps pace with coal’s exit, not the national headlines, which rarely mention the SWIS at all.

– Anjali Rao, Grid & Storage Correspondent

Photo by Valomukitse Arva-Zika on Unsplash