Bayswater and Eraring are running less, not breaking down more

Every quarter AEMO publishes Quarterly Energy Dynamics, its own accounting of what actually happened in the power system, including a table most coverage skips: output, availability and utilisation for each black-coal station in NSW. Availability measures how much of a plant’s capacity was technically ready to run; utilisation measures how much of that available capacity the market actually called on. The gap between those two numbers is where the energy transition stops being a projection and becomes a line in an operating report.

Utilisation of each NSW black-coal station, first quarter of 2025 (open dot) to first quarter of 2026 (filled dot). Redrawn from the data in AEMO’s QED Q1 2026, p. 27 (Figure 37 and accompanying text).

Available, and not called on

Bayswater is the sharpest case in the state. Its availability rose 347 MW, up 15 per cent year on year, the plant offering the market a seventh more capacity than it did last March quarter, and its utilisation still fell ten percentage points, the steepest drop of any NSW black-coal station. Output slipped only slightly, 12 MW on average, because the higher availability partly offset the lower call. Eraring ran harder than any other NSW coal plant, as it usually does, and still eased from 84 to 81 per cent utilisation on availability that was itself slightly up. Across the border of the region, Delta’s Vales Point B was the only NSW station to generate more than a year earlier, and even its utilisation fell, from 73 to 63 per cent, because its availability had risen faster still.

A plant that is broken shows the opposite signature: availability down, utilisation of what remains high. What the March quarter shows, in AEMO’s own numbers, is a fleet that is intact and increasingly left uncalled, the lost running spread across most hours of the day. The distinction matters locally because “the old plants are failing” and “the old plants are being outbid” imply different futures for the people who work in them: the first ends in unplanned breakdowns, the second in scheduled exits like the ones already on the calendar.

What did the displacing

The same report names the competition. Renewables supplied a record 46.5 per cent of NEM generation for a first quarter, up from 42.5; black coal’s share of supply fell from 40.4 to 37.4 per cent; NEM-wide coal generation hit a Q1 record low. The fastest-moving newcomer is storage: grid battery discharge averaged 359 MW, more than triple a year earlier, after 4,445 MW of new battery capacity entered the market in twelve months, the fleet passing 8,000 MW, much of it charging cheaply in the solar middle of the day, exactly the hours coal used to own, and selling into the evening peak. Gas ran at its lowest quarterly average since 1999.

And one number cuts off the easy explanation: demand did not fall. NEM-wide underlying demand set a quarterly record, and NSW’s rose 3 per cent, the only notable increase of any region, with data-centre load up 18 per cent year on year. More electricity was wanted than ever; less of it came from the coal stations that were more available than a year earlier. Wholesale prices tell the same story from a third angle: the NSW quarterly average fell 16 per cent year on year to $73/MWh.

One quarter is a data point, not a destiny, which is why this masthead will re-read the station table every quarter and extend the chart above as the series grows. But the direction of the squeeze is now documented at station level, by name, in the market operator’s own accounting: the Hunter’s coal plants are not being pushed out by their own failures. They are being outbid.

Q2 2026: the gap widened, and every NSW station ran less of what it had

Update, 29 July 2026. We said we would re-read the station table each quarter. AEMO published Quarterly Energy Dynamics Q2 2026 on 24 July, covering 1 April to 30 June, and the squeeze this story described has not eased. Across New South Wales black coal, output fell 213 MW (-4.1 per cent) year on year while availability rose 760 MW (+12 per cent). More plant offered to the market; less of it was taken. AEMO records utilisation falling at every New South Wales black-coal station.

New South Wales black-coal utilisation, Q2 2025 to Q2 2026

Station	Q2 2025	Q2 2026	Output, year on year
Bayswater (AGL)	88%	84%	up 197 MW (+12%)
Eraring (Origin)	76%	75%	down 399 MW (-20%), the largest fall in the state
Vales Point B (Delta)	72%	62%	down 130 MW (-15%)
Mount Piper (EnergyAustralia)	76%	57%	up 120 MW (+21%), on availability up 459 MW (+61%)

Two of those stations generated more power than a year earlier and still ran a smaller share of what they had available, which is the clearest possible statement of what utilisation measures. Mount Piper is the extreme case: it offered 61 per cent more capacity, lifted output by a fifth, and still saw utilisation drop 19 percentage points. Eraring, the state’s largest, went the other way and gave up a fifth of its output.

One thing in our Q1 framing has changed, and it matters. In the March quarter we made a point of the fact that New South Wales demand was rising, so the squeeze could not be explained away as customers simply using less. In the June quarter that reversed: underlying demand in New South Wales fell 104 MW (-1.2 per cent) to average 8,487 MW. AEMO attributes the lower utilisation to both higher availability and reduced operational demand, and we are not going to report only the half that suits the story we told in April. The displacement is still plainly there in the supply mix, but this quarter softer demand is part of the arithmetic too.

The displacement side is emphatic. Coal-fired generation across the market hit a new Q2 low of 13,158 MW (-5.0 per cent), with black coal down 7.3 per cent to 9,322 MW, itself a Q2 record low. Gas recorded its lowest Q2 average since 2003, at 1,050 MW, 30 per cent below Q2 2025. The renewable share of NEM generation reached a new Q2 high of 42.1 per cent, up from 37.1. Grid-scale batteries more than tripled the energy they shift: daytime charging up 1,009 MW (+211 per cent), evening-peak discharge up 1,066 MW (+228 per cent). Market emissions fell to a new Q2 low of 27.1 MtCO2-e (-6.4 per cent). And New South Wales recorded the market’s largest wholesale price fall, down $86/MWh (-53 per cent).

Our view, unchanged and now better evidenced. The Hunter’s coal stations are not failing their way out of the market. They are available, and increasingly not called on, and the reason is visible in the same report: cheaper output is winning the bid, and a battery fleet that barely registered two years ago is now moving enough energy to reshape the evening peak. The one honest caveat is the demand line above. We will read the Q3 2026 edition, due around late October, and report whether demand or displacement is doing more of the work.