Five fixes for the NEM

February 2017

Load shedding in South Australia on Wednesday 8 February and successive ‘close shaves’ in NSW and Queensland as the heatwave spread north have exposed serious weaknesses in the national electricity market (NEM).

For some time now, it has been fashionable to blame South Australia’s near 40 per cent renewable (wind) generation mix for outages in that state. However, reliability problems in regions with much less renewable generation underlines that the challenges facing the power system are much more complex and extend far beyond South Australia.

The NEM’s ‘energy only’ market design means that generators get paid when they are called to run, but not otherwise. It was and remains a fundamentally sound basis for organising generation markets.[1] However, the consequences of some poorly-conceived interventions in that market design are now starting to show.

At the time of its instigation, the bi-partisan renewable energy target (RET) seemed straight-forward: mandate that purchasers of wholesale electricity buy an increasing proportion from renewable sources, and impose stiff penalties for non-compliance. Few appreciated that subsidising one form of generation amounted to imposing a tax on all others.

Compounded by weak demand growth since 2009, the RET has driven a steady exodus of fossil fuel generators from the market. Some aging, CO2 heavy, coal-fired plants had in any case passed their time and needed to go if carbon reduction targets are to be met. But several much newer, CO2 light, gas-fired plants have also been mothballed – squeezed between high domestic gas prices and electricity market revenues deflated by subsidised wind turbines.

These forces have caused the once contemplated role of gas as the ‘transition fuel’ to a much lower carbon generation mix not to materialise, so that we now have a sub-optimal generation portfolio with too few highly reliable, shoulder- and peak-period generators.

With the planned closure of the giant 1600 MW Hazelwood plant next month, market rules and operating protocols must be adapted quickly if the past week’s problems are not to be much worse come the summer of 2017/18.

Here are five suggestions warranting close attention:

1.       The protocols and norms by which market operator AEMO makes decisions need careful review, so that the system is run more conservatively and with greater emphasis on the consequences of ‘getting it wrong’ – for example, in assessing capacity requirements in South Australia, AEMO assumes a contribution towards peak demand from wind of nine per cent, yet when demand was high last Wednesday the weather delivered only half that amount. On its face, a much more cautious approach is needed.

2.       Greater caution as to target levels of supply reliability would also assist the case for AEMO invoking its ‘reliability and emergency reserve trader’ powers more frequently and, longer term, for much stronger interconnection between NEM regions. Heatwaves rarely hit all of the south eastern states simultaneously, and providing for greater diversity of supply across regions is critical.

3.       The generator playing field must be levelled so that the structural disadvantage imposed by the RET on forms of generation that can be scheduled with very high reliability (eg, fossil-fuelled and solar-thermal units) is eliminated, thereby improving the incentives to invest in this form of capacity. There should be no need to throw out the energy-only market design, but there is opportunity for market-saving innovation, perhaps by introducing a price premium for output from scheduled generation, paid for by a discount on the output from semi-scheduled generators (mostly, wind) that only ‘turn up when they can’.

4.       The market price cap should be raised so that it more closely reflects the cost to customers of losing load, and encourages peaking plant to be more available – even if deployed for just a few days per year. For the energy only market design to work as intended, it must not prevent peak period prices from reflecting the high cost of load shedding or non-supply. Ensuring the demand side of the market can see and respond to high price events would also assist energy consumers to make informed choices on peak days.

5.       Lastly, the Victorian and NSW governments could assist by easing their moratoria on gas exploration and development, and facilitating arrangements that enable land-owners to share a greater proportion of the potential value of onshore gas reserves. Improved availability would take the pressure off domestic gas prices and strengthen the viability of gas-fired generation.

The first four of these measures would all add to the cost of generating and delivering electricity, and so the prices paid by consumers, while the fifth is also far from popular. However, the cost to consumers and the economy as a whole of an unreliable power system is a much higher price to pay. Decarbonising our energy system involves hard choices, and time is running out for those choices to be made.  


[1] By contrast, generators in Western Australia (not part of the eastern states market) get paid just for being available, and WA consumers now pay a high premium for capacity they do not need.


Greg Houston was interviewed in relation to the challenges facing the NEM on ABC’s 7.30, Monday 13 February. See:

Article originally published on LinkedIn
Posted on February 14, 2017 and filed under Articles.