$16,600 a megawatt-hour, then minus $57 a megawatt-hour, inside the same trading day. That’s not a glitch in AEMO’s dispatch engine. That’s South Australia on a windy, sunny afternoon followed by a still evening, and it happened more than once this year. Wholesale electricity price volatility in the NEM has become the thing everyone complains about and almost nobody explains properly, so let’s have a go.
Start with the basic fact that trips up most commentary on this. The National Electricity Market was never built to produce a stable price. It was built to produce the right price, every five minutes, for whatever the system needs at that instant. Stability was never the design brief. Reflecting true scarcity and true abundance, second by second, was.
The five-minute mechanism nobody outside the industry thinks about #
Every five minutes, generators bid a price and a quantity into AEMO’s dispatch engine, and the engine picks the cheapest combination that meets demand. I’ve written before about how the NEM actually dispatches power every five minutes, and it’s worth restating the core point here: the price that clears the market is set by the last unit of generation needed, not the average cost of everything running.
That’s called marginal pricing, and it’s standard in electricity markets worldwide. But it has a brutal consequence in a grid with lots of near-zero-marginal-cost renewables sitting next to expensive, inflexible gas and hydro peakers. When rooftop and utility-scale solar flood the middle of the day, the marginal unit needed might cost almost nothing. When the sun drops and a gas peaker has to switch on to cover the evening ramp, the marginal unit might cost hundreds of dollars a megawatt-hour. Same market, same day, two completely different physics.
Negative prices are becoming routine, and that’s the real story #
Forget the headline spike for a second. The more structurally important trend is how often prices go negative in the middle of the day, particularly in South Australia and increasingly in Victoria and NSW. AEMO’s own quarterly energy dynamics reports have tracked rising negative-price intervals for several years running, and the pattern keeps extending. I covered the mechanics of this in rooftop solar and the midday minimum demand problem – there’s now so much rooftop solar pushing into the grid at midday that utility-scale generators sometimes pay the market to keep running rather than switch off and risk losing synchronisation or contractual position.
That’s the quiet half of the volatility story. The loud half is the evening price spike, when solar falls away, demand climbs as people get home, and whatever’s left to fill the gap (batteries, gas, hydro, interconnector flows) gets paid scarcity prices for doing it. Put the two together and you get a daily price curve that looks less like a gentle wave and more like a seismograph during an earthquake.
South Australia is the pattern, not the exception #
I keep coming back to South Australia because it’s where the rest of the NEM is heading, several years early. I’ve written at length about South Australia as the grid that keeps teaching the NEM lessons, and the price volatility there is the clearest version of the phenomenon. On a good wind-and-solar day the state can run close to, or briefly above, 100 per cent instantaneous renewable supply. Prices often collapse to zero or below. Then conditions change – a wind lull, cloud cover, a plant trip, and the state leans hard on gas and the Heywood and Murraylink interconnectors, and prices can rocket.
The state government and AEMO have both pushed grid-forming inverter technology and battery capacity specifically to dampen this swing, which is a tacit admission that volatility isn’t just an inconvenience. It’s becoming a system stability question too. I’ve covered why grid-forming inverters are suddenly something the grid needs, and South Australia’s experience is Exhibit A for why that technology matters more than the marketing material suggests.
Who actually profits from the swings #
Here’s where I’ll push back on the usual framing. A lot of retail commentary treats volatility as purely a consumer problem, and for anyone on a flat tariff with no hedge, it is. But volatility is also a deliberate profit mechanism for anyone holding flexible, dispatchable capacity. Battery operators buy at the negative or near-zero midday price and sell into the evening peak. That’s not a side effect. That’s the entire business case for grid-scale storage in this market.
Look at the economics behind projects like the Waratah Super Battery or the fleet Akaysha Energy has been building across the NEM: the revenue model depends on price spread, not price stability. A flatter wholesale price curve would actually make batteries less profitable, not more. So when politicians talk about “smoothing” volatility as a universal good, I reckon they’re missing that the storage build-out everyone wants depends on some volatility surviving. You can dampen the extremes without eliminating the spread, but you can’t have it both ways entirely.
Retailers, hedging, and why your bill doesn’t swing like the spot price #
Most households never see this volatility directly because retailers hedge against it, buying futures contracts, owning their own generation, or both. That’s the gentailer model in a sentence, and it’s why I’ve spent time comparing AGL versus Origin as two gentailer strategies – both companies effectively run internal insurance books against the exact price swings this article is describing. When the hedge book is well matched to the retail load, spot volatility is a non-event for the company’s earnings. When it isn’t, as happened to several smaller retailers during the 2022 price crisis triggered by coal outages and the east-coast gas squeeze, the mismatch can be existential.
This is also why the AER’s Default Market Offer exists as a backstop for customers without hedging sophistication of their own, and why the Capacity Investment Scheme tender process, which I’ve gone through in how the Capacity Investment Scheme actually works – has become the government’s preferred lever for locking in new dispatchable and storage capacity ahead of need, rather than waiting for the spot market alone to signal it. The CIS is essentially a bet that private capital won’t build enough firming capacity fast enough purely off volatile spot revenue, so the Commonwealth is underwriting the gap.
Coal’s exit is accelerating the swings, not smoothing them #
There’s a comforting theory that once enough coal retires and enough renewables and storage arrive, the price curve settles down. I’m sceptical of the timeline on that. In the near term, each coal unit that exits removes a large, relatively predictable block of supply and replaces the gap with a more weather-dependent, more price-sensitive mix. The pace mismatch is real – I’ve tracked it in coal exits versus replacement builds, and until storage and transmission catch up fully, every major coal closure is a volatility event in its own right, not a calming one. Eraring’s extended life is itself a hedge against exactly this risk, buying time for Waratah and other batteries to firm up before the next big block comes off.
What AEMO and the AEMC are actually doing about it #
The market bodies aren’t ignoring this. AEMO’s Integrated System Plan explicitly models the need for firming capacity to manage this exact volatility pattern at scale, and the AEMC has run several rule-change processes looking at how dispatch and pricing rules interact with batteries and fast-start gas. The regulatory settings – things like the administered price cap, currently set well into five figures per megawatt-hour, and the market floor price, set deeply negative – are themselves an acknowledgement that extreme swings are a feature the market is designed to tolerate, within limits, rather than suppress. You can read AEMO’s own quarterly energy dynamics data directly on their website if you want to see just how often the extremes have been hit this year, and the AER publishes separate wholesale market performance reports that track the same thing from a compliance angle.
A small cricket aside, because I can’t resist it: watching a price chart swing from a few dollars to five figures and back inside an hour reminds me of a T20 run chase where the required rate swings wildly over after over – ugly to watch if you want a tidy story, but it’s exactly the scoreboard pressure that’s meant to produce a result. Nobody designs a run chase to look smooth. Nobody should expect this market to either.
The honest shape of where this goes #
My reading is that wholesale price volatility in the NEM gets more pronounced before it gets calmer, as more rooftop and grid-scale solar push midday prices toward and below zero, and as coal retirements remove dispatchable supply faster than transmission and storage can fully backfill it. The spread between cheap and expensive hours is the signal that’s supposed to bring batteries, pumped hydro and demand response into the market at the right pace. Projects like Borumba pumped hydro in Queensland are effectively multi-decade bets on that spread persisting long enough to repay the capital.
The uncomfortable truth for anyone wanting a tidy, stable wholesale price is that stability and the transition don’t point the same direction right now. A genuinely smooth price curve would mean either far more firming capacity than currently exists, or a slower renewables build than the Integrated System Plan assumes. We don’t have the first yet, and nobody’s seriously proposing the second. So the swings stay, and the real question for policymakers isn’t how to eliminate them. It’s how to make sure the households and small businesses without a hedge aren’t the ones left holding the bag when the evening price spikes again.
– Marcus Wren, Editor
Image: Queensland State Archives from Runcorn, Queensland, Australi (Public domain) via Wikimedia Commons