Gas in the NEM: Is there a case for a new and expanded RERT?

Gas powered generation (GPG) will play a critical role in the decarbonisation of the National Electricity Market (NEM) as it is the only currently viable technology that can facilitate the operation of a grid with weather dependent variable renewable energy (VRE) and storage after coal has exited the market. This fact is accepted by all governments, market bodies and participants. The services that GPG will provide are as follows:

1. Firming weather dependent VRE and storage. This could involve managing solar output profiles beyond what batteries can support or managing business as usual short duration VRE droughts;

2. Insurance to support the system during High Impact Low Probability (HILP) extended VRE droughts; and

3. System support through provision of essential system services (ESS).

Figure 1 illustrates AEMO’s forecast role for GPG in the NEM.

Figure 1: Actual and forecast GPG capacity (MW) and capacity factors

Source: AEMO 2024 Forecasting Assumptions Update Workbook.

As can be seen from Figure 1, from 2023 AEMO forecasts the GPG generation mix moves to predominantly peaking plant by 2040 and the expected capacity factor from peaking plant increases from a little over 2.4 per cent to almost 8 per cent. Based on the retirement dates in the data and forecasts, almost 8 GW of new GPG will be required by 2040.

Investment

For the transition to a high VRE energy system to proceed, substantial investment in GPG and associated infrastructure must occur. However, from the investor’s perspective there is too much unmanageable risk to support a business case and obtain finance.

Mechanisms which would support capacity, including GPG have been considered most recently by the Energy Security Board, who advocated for a technology neutral approach.  However, due to jurisdictional objections, this mechanism eventually morphed into the Capacity Investment Scheme (CIS) which rather than incentivise GPG (and long duration storage) investment, actively excluded GPG and instead expanded it to include VRE. Hence, the CIS is bringing forward significant supply into the market that in the absence of a CIS would suffer revenue adequacy issues. The impact of this ‘bring forward’ of VRE on both new and existing firming generation was highlighted by the AEC’s submission to the Expert Panel, which presented modelling from Endgame Analytics.[1] 

In short, Endgame Analytics found the ‘bring forward’ of VRE depressed spot prices and created revenue adequacy issues for both incumbent and new entrant GPG even though it is what the market requires in the medium to long term.

Markets and the NEM Wholesale Market Settings Review

The NEM wholesale market settings are currently being reviewed by an Expert Panel and one of the key issues they are grappling with is how to ensure necessary GPG investment eventuates in a timely manner.[i] They are trying to solve for two very different types of GPG investment, firming and insurance (for HILP events ie, extended VRE droughts), and they are seeking a market-based mechanism. Figure 2 illustrates how bifurcated this market is. Firming GPG is represented by the horizontal part of the curve and insurance by the vertical part.

Figure 2: LCOE ($/MWh) at node and capacity factors for new 400 MW OCGT

Source: AEC analysis

If the approach to ensuring these two services will be available assumes one solution fits all, it will result in a dysfunctional market and will be excessively expensive because the insurance part of the market will create an oversupply in the firming market. Ifleft to its own devices, the market will not support the insurance capacity such that we only have a firming market. Hence, two separate markets are required. An amount of GPG firming capacity that operates as a market with some level of underwriting and an insurance capacity that is excluded from the market unless required to perform reliability emergency reserve trader (RERT) obligations.  

New expanded RERT

The greatest risk facing the future NEM is extended VRE droughts. In these situations, storage and GPG will be placed under immense pressure to the point where GPG is required to charge storage and the gas system is unable to fully fuel GPG such that it runs on diesel as well. In previous papers we have demonstrated there would be load shedding well in excess of the reliability standard, which is 0.002% unserved energy (USE).[ii] Events like this would cause damage to the economy and a loss of confidence in a high VRE NEM. Based on our modelling, at least 5 GW of additional GPG capacity is required to ensure reliability.

With respect to weather induced HILP events, we cannot be sure that these events remain low probability in the future. This is because the climate is changing which means confidence in statistics and probabilities derived from historical observations need to be discounted.

While the two types of GPG services use the same technology, they are two distinctly different services which require different cost recovery methods. The former operates relatively regularly and will offer hedging products for retailers. In contrast, the latter has very low expected utilisation and requires extremely high spot (or contract) prices to recover its capital costs. It could offer hedging products for HILP events, but it is uncertain whether participants would be able to afford these products and whether they should. Managing this type of risk is likely to be beyond the scope of the market and another approach is necessary.

Hence, they should be treated as two different types of market participant that supply two distinct product markets. Comparisons with other markets:

• The general insurance market that can handle most claims from its own resources and the latter is the monoline reinsurance market where general insurers lay off large amounts of risk that exceeds their own funding envelopes. Reinsurers can also lay off some of their risk by issuing catastrophe bonds where purchasers receive a high yield. Nevertheless, governments cover any residual risk.
• The RBA’s role in providing both liquidity and asset purchases to support the system during financial crises. This is a product that is beyond the scope of financial markets because they can’t issue currency. And most importantly the RBA only does this in a crisis and does not distort the market which functions well in ‘normal’ times.
• Multibillion dollar state government investments in desalination and potable recycled water plants that are catchment precipitation independent to mitigate HILP drought events.

In the first instance, insurance GPG must only be allowed to generate to prevent USE. If it is allowed to operate in market it will increase the MWs of capacity in the competitive firming market thereby creating oversupply and make those market participants uneconomic. Insurance GPG should be funded differently, earning a return and costs recovered irrespective of whether the service operates.

The insurance service would be provided under new reformed RERT arrangements. This means it would sit outside the market and not require a market solution to fund it. Rather it could be funded through economic regulation similar to how networks are currently regulated. Alternatively, it could be less intrusive regulation where tenderers or bidders define how they are funded as part of a competitive process. This approach would enable firming GPG to eventually enter the market knowing it is not going to be crowded out by insurance GPG. Either way, investors in RERT assets would require a lower rate of return relative to firming GPG, because their required cash flows are guaranteed under regulation. Initial modelling indicates almost a $600 million reduction in the net present cost of commissioning 5 GW of RERT GPG assets, if a regulated cost of capital is used instead of a market one.[iii]

To put these costs into perspective some relevant metrics for NEM Network Service Providers (NSPs) for FY23:[2]

• $12.5 billion revenue/year ($1,144/customer)
• $6.8 billion/year capex
• $116 billion total regulated asset base (RAB, $26.1 billion TNSPs and $89.9 billion DNSPs)
• $10,609 RAB/customer

AEC modelling has indicated that at least an additional 5 GW of GPG is necessary to mitigate VRE drought risk in 2040. The capital cost of this would be approximately $6-7 billion. At a 5-6 per cent nominal vanilla WACC, the annual return on asset would be $300-420 million. Fixed operations and maintenance (FOM) at $10,000/MW produces a total FOM of $50 million.[iv] Hence, when compared to costs and RABs associated with the NEM’s network infrastructure, they represent a prudent investment. The capital (‘RAB’) per customer would be $520 - $600 and annual revenue per customer would be $30-40. For households this would be lower because larger users would pay more.

Furthermore, if the RERT is dispatched it will be bid in at Market Price Cap (MPC) plus $100. This ensures non-RERT generators would not be crowded out. Settlement would still be at the MPC. Any spot revenue earned could be used to defray the regulated revenues paid to RERT owners.

Conclusion

It is clear a level of support is required to incentivise investment in firming GPG capacity. Failure to do so will slow the transition and keep coal in the market longer than is necessary. Secondly, it is critical the HILP extended VRE drought risk is mitigated with additional GPG capacity otherwise state governments will continue to retain coal until it is proven that it is not required. Finally, the most cost effective and efficient way to achieve this is through a new expanded RERT that benefits from the low cost of capital for investors in a regulated asset.

This paper only seeks to introduce this approach into the current discussion and any modelling results are a first pass. If this is to be implemented, many details including governance and design issues will need to be addressed to create a new expanded RERT.

[1] https://www.energycouncil.com.au/media/syxld2km/20250214-nem-whs-mkt-set-review-initial-consult-aec-sub-final.pdf

[2] AER State of the energy market 2024 – Electricity networks.

https://www.aer.gov.au/system/files/2024-11/State%20of%20the%20energy%20market%202024.pdf

[i] https://www.dcceew.gov.au/energy/markets/nem-wms-review#dcceew-main

[ii] https://aemo.com.au/-/media/files/stakeholder_consultation/consultations/nem-consultations/2024/2026-isp-methodology/submissions/australian-energy-council.pdf?la=en

[iii] Market pre-tax WACC of 12 per cent produces NPC of $6.7 billion and regulated pre-tax WACC of 7.6 per cent produces NPC of $6.1 billion.

[iv] AEMO 2024 Forecasting Assumptions Update Workbook.