Minimum demand season: a killing field for old coal

Welcome to springtime, or as it’s known by the electricity industry, minimum demand season. Every spring the National Electricity Market returns with megawatts of new solar and wind. The resulting mega-squeezing of other generators is what helps put the oldest and frailest coal power stations out of business.

That would be a great lower emissions outcome if anybody was building the hundreds of megawatts of lower emissions firming capacity needed to fill in when renewables aren’t generating. But no one is.

The original test case for the impacts of big renewables driving minimum demand was South Australia. Chronic minimum demand events there kept forcing the two units at the brown coal Northern power station to switch off, which is what ultimately killed it.

The next will be Victoria and Yallourn. Victoria lost around 22 per cent of its firm generation when Hazelwood power station closed in 2017. Yallourn closing would take that down to 45 per cent. That’s a really big hole in the state’s electricity supply. The Electricity Statement of Opportunities currently assumes this will close in 2029-30. But what if it was forced out earlier?

The team at WattClarity observed that scheduled demand in Victoria reached a new record daytime minimum of 2,992 MW at lunchtime on Saturday 29 August 2020. This record was then shattered 8 days later on Fathers Day (September 6 2020), down to 2,690MW at lunchtime. Scheduled demand is the demand met by fully dispatchable supply like coal, hydro, gas and batteries. What is even more remarkable abut this event is that it includes the full operation of the Portland aluminium smelter (around a 600MW load).

Northern and South Australia

Coal fired power stations are designed to switch on and stay on for long periods of time. They take time to start and generate within a specified operating range once on. turning them off and on rapidly accelerates their operational decline, fatiguing high temperature alloys used in the boilers.

As renewable generation increased rapidly in South Australia, the owners of Northern Power station, Alinta, faced either having to generate at a loss for days or weeks at a time during lower demand seasons like spring and  autumn, or be forced to turn off and on. They attempted, unsuccessfully, to contract directly with enough customers to be more immune from the market price, but without success.

The death rattle for coal is the rate at which units are switched on and off. Stable, high output units in generators like Stanwell in Queensland might only switch off once or twice a year. Generally staying in the lower end of single digits is desirable. Higher numbers either indicates age requiring more content maintenance (like Hazelwood at the end of its operational life, Gladstone now). Or it can be forced by market conditions.

In 2013 Northern’s first unit was off 14 times, 16 times in 2014 and 19 times in 2015. The second unit was off 8 times, 8 times and 22 times. By 2015 Alinta was switching its two units off and back on 41 times a year. That’s a lot. It looks like this.

Northern power station generation per day, July 2014-16 Source: NEM Watch

Alinta announced it was closing the power station in September 2015 because it could not profitably operate with 41 per cent of generation in South Australia coming from wind and solar.

Yallourn and Victoria

Yallourn is the oldest and the most commercially vulnerable of the three remaining brown coal generators left in Victoria. It has four 400MW units totalling 1600MW of capacity. The minimum output for each unit is around 250MW.

Yallourn is really two power stations joined together: units 1 and 2 were commissioned in 1975, units 3 and 4 in 1982. The two younger units have generally been more reliable and left on longer, while, according to Watt Clarity’s Generator Report Card, units 1 and 2 were re-started 25 and 22 times respectively in 2018.

Managing Yallourn is already an operational dance for EnergyAustralia. Older units need more constant maintenance, all four units need to be on for peak demand season in summer and winter and more units out of service in low demand spring and autumn. The generation mix will reflect EnergyAustralia’s contract position in the market. Like right now.

The last year of Yallourn operation looks like this:

Yallourn Power Station generation by day September 2019-20 Source: NEM Watch

It’s a bit tangled, but it is easy to see how all four units were operating during summer. The overall busyness is less than ideal for and old coal generator. That suggests either weak market conditions or fragile reliability. Or both.

Ironically Energy Australia had 3 units on during the two recent low demand weekends in Victoria. The three operating units crunched down to their minimum output for a few hours on both days, with around 800 MW of surplus electricity transmitted into Tasmania and NSW. Yallourn was generating around 750MW of electricity at negative spot prices for most of the day – down to around -$40/MWh during the middle of the day.

EnergyAustralia won’t be badly out of pocket from Fathers Day because it will have contracted a lot of its supply to retailers to hedge against that price risk. But as those sorts of price signals become more enduring, as they are, they impact future contract prices and the cost of hedging that risk.

Victoria sourced around 18 per cent of its electricity last year from wind and solar. It’s still got a way to go to reach South Australian levels. But if the financially distressed Alcoa smelter at Portland closes some time this decade and renewables keep increasing, then the maths of Yallourn remaining commercially viable shortens considerably.