Too low for zero: I’m on a losing streak

On the morning of 6 February 2020 AEMO issued market notice 73857 in response to the enduring islanding of the SA network informing the market that when SA demand falls under 800MW a number of generators, including more than 500MW of wind and 300MW of solar will be constrained down to zero.

The periods of lowest demand in SA typically occurs during the middle of sunny mild days (September – January), when rooftop solar is beginning to ramp up their output but the weather is not yet hot enough to require air-conditioning. Since 2017, 96%-98% of all settlement periods where demand was less than 800MW in SA occurred between 10am and 5pm. The number instances of sub 800MW periods of demand in SA is increasing exponentially with the number of periods increasing from just under 400 to more than 850 in 2019 compared to 2018.
Incredibly in the first month of 2020, SA saw 184 instances of sub 800MW demand (that is more that the total number of times for 2016) and four times more than January 2019. Average max temperatures in SA were almost 4 degrees lower in January 2020 compared to 2019 assisting to keep demand lower.

Historically February has only seen a handful of low demand periods (avg. 10 periods since 2017) and therefore generators would not necessarily expect to be curtailed often between now and the beginning of September 2020, depending on how long the interconnector takes to get repaired. However, given the sharp rise in the number of low demand periods in January, we may see more significant increases in February if the milder weather holds.

If generator outputs are going to be curtailed to zero during these periods then what is the financial cost if they were exposed to the spot price? Last year, on average, the listed wind and solar farms were cumulatively generating ~146MW and 139MW respectively. The average spot price when demand has been less than 800MW in February is only ~$27/MWh. What is interesting when looking at the generation profiles of the wind and solar assets during these low demand periods is that wind is unresponsive to negative priced periods. The total spot price revenues (if they were spot-exposed for all their output) of the energy generated during the low demand periods by the identified wind generators (Hornsdale wind farms, Lincoln Gap and Willogoleche) was -$1.7M in 2019 compared to Solar which was $1.9M (difference of $3.6M).

This is a result of the contracts they have been able to secure, with all stages of the Hornsdale wind farms contracted with the ACT government which pays generators during negative pricing, compared to the newer solar farms which are less inclined to generate during high priced periods particularly Tailem Bend.

The real issue is whether these curtailments will persist long-term in the market as AEMO tests participants’ responses to this market intervention. It remains to be seen whether this method for rectifying system issues as an alternative to developing market mechanisms such as ‘demand turn up’ or utilising larger portions of lower FCAS (localised in SA for this purpose). Given the current market structure, it appears that in the short term price-hedged renewable players will be the most at risk during these curtailment periods (e.g Hornsdale).