On the 25th of May at 1:45pm  Queensland Coal and gas were generating about 5,300 MW out of a total of 7,500 MW. Then Unit 2 at Callide C exploded taking 400 MW of generation off the Queensland grid. Immediately its twin also disconnected, whether due to damage or safety reasons is not clear.  However the total loss was still only 800 MW or just over 10% of Queensland supply at the time. Over the next few minutes Queensland swapped from exporting about 400 MW to importing 200 MW so theoretically recovereing 2/3rds of the lost generation

However twenty minutes later another nearby unit, in a different building went offline, together with a transmission line and at least another 7 coal and gas generators. Then due to lack of inertia, wind and solar plants in North Queensland were immediately limited to 50% of peak (not actual) output. This resulted in 400,000 customers temporarily losing power.  According to some expert observers AEMO did a sterling job preventing the state going black. Probably having learnt some lessons from SA.

Now although Queensland has 12.8 GW of dispatchable capacity of which probably 11 GW was available after allowing for the first Callide outage and maintenance, it ramps up so slowly that it could not meet 4.6 GW of net demand (after allowing for wind and solar generation and swap from exports to imports at the time). In fact the coal plant disruption effectively managed to drive half the wind and solar offline as well.

Since recovery, spot power prices have remained high, averaging $296/MWh in the 21 days since the incident. At the same time last year they were averaging $39. Almost every evening QLD prices have peaked to between $2,500 and $14,000.  These price effects spread throughout the grid as Queensland exported less and imported more power, although still remaining a net exporter, For example NSW has averaged $196/MWh so far in June vs $48 last June even though renewable output is up significantly. Victoria is up from $48 to $65

Contrast the fallout from this event with the relatively mild effects of the flood at Yallourn in Victoria. Yallourn output on last Saturday at 2PM was about 700 MW which was then ramped down to 200MW as the threat of a mine flood became apparent. That was 14% of Victoria’s fossil fuel generation or about 10% of total generation at the time. It also corresponded to a period of very low wind. Yet there were no blackouts and prices only moved up briefly. Since then average prices only reaching $100 and peak prices rarely went above $300 in spite of still, overcast conditions and continuing lack of output from Yallourn.

Now a key difference is that Victoria ramped down gradually instead of in two big steps so one would have expected a more benign response. However Queensland has only been running its remaining 6.6 GW of coal and 1.2GW of base load gas plants at 65% and 45% of capacity respectively, with maximum coal output of 4.9 GW. Seven of its 22 coal generators have been offline for at least the last week, so either the remaining coal plants are off line for service and are nowhere near the 24/7 availability that their proponents claim, and therefore can’t be brought online or, the owners are withdrawing capacity from the market to force prices up. Such behaviour is more difficult in Victoria because renewables supply (30%) almost double Queensland’s share of 18% and Victoria has a theoretical maximum import capacity of about 2,600 MW vs Queensland’s 450 MW. 

If Queensland had followed SA and Victorian practice of installing batteries, somewhere between 300 and 600 MW of batteries would have supplied sufficient reserves to limit voltage and frequency instability so the second trip would almost certainly not have happened. This week, with 7 coal units offline and wind and solar still limited, Queensland manages to export power every day. Therefore it has plenty of capacity to recharge batteries and pumped hydro to meet the evening peak without requiring huge jumps in prices. Even if batteries had only reduced the jump in spot prices by one third and spot prices only account for 15% of Queensland power trade, sufficient batteries would have saved Queensland and NSW customers $60-100m in the last three weeks, not including the costs of the blackouts. Almost every day they would reduce peak prices and improve the operating efficiency of remaining coal and gas by absorbing power when demand is low and releasing it when demand is high. Alternatively, additional coal and gas plants would only add to the surplus capacity and do little or nothing to improve grid reliability in a case such as this.

The superiority of renewables/gas /storage has been clearly demonstrated in SA where prices this financial year are down 62% since the peak year in 18/19, whereas NSW is only down about 20%. The SA renewable/gas grid is not only cheaper to run but has at least twice survived generation/transmission losses of 30%+ of load, a feat totally unimaginable in Queensland or NSW.

It is not as though Queensland has not had warnings, on 9th of October 2019 Kogan Creek coal power plant tripped and caused small blackouts in southern Queensland and NSW. On 25th of August 2018 a lightning strike on the 800 MW inter-connector between NSW and Queensland caused short localised blackouts in Queensland, NSW and Victoria, and almost caused the entire NSW system to go black.

A system black in NSW in the middle of the grid is a far more serious issue than the last one in SA. It could quite easily have cascaded into Victoria and because of NSW reliance on coal, it would not have recovered in 6 hours like most of SA, it would have taken days and maybe weeks. The Texas power authorities estimated that it would have taken months if their system had gone offline in February to bring everyone back and Texas has a far more flexible generating fleet than NSW.

In conclusion, both reliability and economic competitiveness of coal fired power stations has been vastly exaggerated by their proponents and even if there were no such consideration as pollution, water use etc. a renewable grid backed by gas and storage is a a far more reliable and economical system at today’s state of the technology.