Why we need a new NEM: AGL

There is no magic button to push which will integrate climate and energy policy, writes AGL’s Tim Nelson. It’s time to redesign the NEM.


The national electricity market (NEM) was created in the 1990s as an “energy-only” market in which electricity generators are paid for the energy they produce but not the capacity they make available.

Most of the time, excess capacity in the market ensures that prices reflect short-run marginal costs (that is, the cost of buying coal and gas). But at times of peak demand prices are permitted to increase significantly, which allows generators to recover the heavy fixed costs associated with building the required infrastructure. When the NEM was created, there was little focus on reducing greenhouse gas emissions and almost no wind or solar generation in operation.

At the turn of the century, things began to change. Concerns about climate change resulted in a number of policies being introduced to incentivise investment in new renewable electricity generation capacity. The 20% Renewable Energy Target and various state government premium feed-in tariff policies resulted in relatively rapid investment in renewable generation. But almost no attention was paid to the interaction of these climate change policy decisions with design of the energy-only NEM.



There has been significant investment in new electricity generation capacity since the NEM was created in the 1990s. However, more than half of this investment has been driven by climate change and renewable policies rather than NEM wholesale market pricing. Electricity demand has also fallen relative to expectations. These factors have led to wholesale market pricing being insufficient to recover the long-run marginal costs of an optimal plant mix (being the optimal generation fleet for meeting demand).

In theory, ageing emissions-intensive power stations that are surplus to requirements would be permanently retired and decommissioned. However, there are four potential barriers to exit for incumbent power stations that impact on decision making. Firstly, existing plants have very low economic costs and so are likely to “sweat” assets until the marginal cost of operations exceeds revenues obtained from reduced operation.

Secondly, participants have a disincentive to exit due to first-mover disadvantage. Third, site remediation costs are often significant and deferring closure reduces costs by putting them off into the future. Fourth, policy uncertainty (a significant feature of Australian energy markets) impacts on decision making.

If power stations are not decommissioned due to these potential barriers to exit, their capacity remains available for dispatch with advanced notice. The presence of mothballed plant capacity depresses expected future wholesale electricity prices despite not physically generating electricity because it can be recalled with notice. Therefore, potential barriers to exit may result in mothballing and reduced operating duties rather than permanent closure.



Around three-quarters of the NEM’s thermal (coal and gas) plant are older than their original design life. These plants will need to be replaced with low emissions capacity (such as renewables, gas and advanced batteries) over the coming decades to ensure system reliability is maintained. However, potential barriers to exit and relatively stagnant electricity demand are likely to supress wholesale electricity revenues below the level required for new investment. This is a problem for investment in both renewable and complementary capacity (such as open-cycle gas turbines). Renewable generators will be increasingly reliant upon policy-based subsidies rather than market revenue. This is likely to deter investment.

Ageing emissions-intensive power stations will close eventually. However, due to the lumpy nature of capital investment and the potential for closure announcements to be made in a way that does not allow new capacity to be constructed before closure occurs, pricing volatility may become extreme. There is a role for governments to establish policy that facilitates orderly rather than disorderly exit of emissions-intensive aged power stations. Such policy could be based upon age (such as a Canadian rule which requires power stations to be closed or retrofitted with carbon capture and storage when they turn 50), emissions intensity or a market mechanism (as proposed by Jotzo and Mazouz).



The economic characteristics of solar PV and wind are slightly different to those of coal- and gas-fired power stations. While solar PV and wind farms involve significant capital costs, they have very low short-run costs (the sun and wind are free!). Coal- and gas-fired power stations have higher short-run costs as they must purchase gas and coal to produce electricity. Therefore, existing solar PV and wind farms will generally always be more cost-effective than existing coal- and gas-fired power stations if the wind is blowing or the sun is shining.

But solar PV and wind generation are considered non-firm or intermittent. In other words, if the sun isn’t shining and the wind isn’t blowing, available capacity doesn’t produce any energy. There is a need for other “firm” capacity to be available to meet demand when wind and solar PV aren’t available. But this capacity is only remunerated when it is needed via the energy it produces.

Pricing volatility is effectively the economic means by which energy-only markets provide revenue adequacy for an optimal generation mix. Heavy fixed costs associated with building power stations are recovered at times of peak electricity demand. But with the introduction of very low short-run cost renewable generation via climate-related public policies such as the 20% Renewable Energy Target, pricing volatility must become extreme to ensure capital costs of firm complementary thermal generation or battery storage technologies can be recovered.

However, extreme pricing volatility, which is required for an energy-only market with high penetration of renewables, is inconsistent with real-world constraints and community expectations. A 2016 study found that the NEM would require a market price cap of between $60,000 and $80,000 per MWh for revenue adequacy if the system was supplied by 100% renewable energy – six times higher than today. Prices would need to be able to increase by a factor of around 1,500 in half an hour.

It is not clear why such an outcome would be desirable given the high known fixed-cost nature of renewable energy. In simple terms, it is hard to justify continued reliance upon an energy-only market in an environment where the cost of energy is zero but the comparative cost of developing capacity to produce the energy is high and fixed.

Continued use of an energy-only market while pursuing high proportions of renewable energy for climate change-related public policy purposes is likely to be unacceptable to generation financiers, retailers, customers and governments.

In the long-term, it is almost inconceivable that banks would provide finance to new projects if costs are largely fixed but revenues can fluctuate 80,000% in half an hour. As such, reconsideration of the NEM’s energy-only market design appears inevitable.



The current 20% Renewable Energy Target requires investment in around 5GW of new renewable capacity by 2020. However, the RET legislation expires in 2030. For investments made in coming years this may be problematic as large-scale renewable generation investment cases are based upon 15 or more years of revenue.

External to the RET, the Queensland government has committed to a 50% renewable energy target in 2030 and the Victorian government is aiming to ensure 40% of electricity is produced from renewables by 2025. The policy frameworks underpinning these commitments are still being developed. It is important that policymakers consider how to develop nationally consistent policies for incentivising renewable energy investment.



So where does all this leave the NEM? In 2015, Australia committed to reducing greenhouse gas emissions by 26-28% relative to 2005 levels by 2030. It is highly likely this will require significant deployment of new renewable electricity generation across the NEM. There are three main policies that require attention.

First, it is necessary to ensure that the transition to a low-emissions electricity system occurs in an orderly rather than disorderly way. A policy framework that ensures the orderly retirement of ageing emissions-intensive power stations is required. Such a policy has precedent given the existence of age-based limitations on power station operations in Canada.

Second, it is necessary to revisit the design of the NEM. Energy-only markets require extreme pricing volatility to produce adequate revenues to incentivise new investment in renewables and complementary capacity (such as open-cycle gas turbines or advanced batteries). A new market design that values firm capacity as well as energy should be developed. Importantly, a traditional capacity market is unlikely to be the answer. Rather, new thinking will be required to develop a robust market design that facilitates the transition to a decarbonised energy system.

And third, it is necessary to develop robust and nationally consistent policy frameworks for incentivising new renewable generation capacity.

The current trajectory of climate and electricity policy puts Australia on a long-term collision course with customer expectations in relation to stable pricing outcomes being unmet.

It is imperative that Australian policymakers better integrate climate change and energy policy. Such integration will allow Australia to better navigate the policy trilemma of competitively priced energy, reduced greenhouse gas emissions and energy security and reliability.


Tim Nelson is the head of economics, policy and sustainability at AGL Energy and an adjunct associate professor at Griffith University.

Send this to a friend