Case study: Peaker plant replacement & capacity markets — a city or utility pilot and the results so far

South Australia's Hornsdale Power Reserve, the battery system originally dubbed the "Tesla Big Battery," has displaced an estimated 35% of gas peaker plant dispatches in the state since reaching its expanded 194 MW / 259 MWh capacity in late 2020, while cutting frequency regulation costs by over A$150 million in its first six years of operation (Australian Energy Market Operator, 2025). What began as a A$91 million bet by Neoen and Tesla following a catastrophic statewide blackout in September 2016 has become the most studied utility-scale battery project in the Asia-Pacific region and a reference model for peaker plant replacement strategies worldwide. As of early 2026, the project has been joined by a network of additional battery installations across South Australia totaling over 900 MW, collectively reshaping how the National Electricity Market handles peak demand and grid stability.

Why It Matters

Peaker plants, typically open-cycle gas turbines or older diesel generators that operate only during periods of high electricity demand, represent some of the most carbon-intensive and expensive generating assets on any grid. Globally, peaker plants run for fewer than 1,000 hours per year on average but can account for 10 to 20% of total electricity-sector emissions in systems where they are prevalent. In Australia's National Electricity Market (NEM), gas peakers historically set wholesale electricity prices during the top 5% of demand hours, with spot prices during these periods reaching A$5,000 to A$16,600 per MWh, the market's price cap (Australian Energy Regulator, 2024).

The economics of peaker replacement have shifted decisively. Lithium-ion battery storage costs fell from approximately US$1,200 per kWh in 2010 to US$139 per kWh in 2024, a decline of 88% (BloombergNEF, 2025). For systems designed to dispatch for 1 to 4 hours, battery storage now delivers a lower levelized cost of peaking capacity than new-build gas turbines in most markets across the Asia-Pacific region. The Australian Energy Market Commission's 2024 Reliability Panel review confirmed that batteries and demand response together could replace 2.4 GW of retiring thermal peaking capacity across the NEM by 2030 without compromising reliability standards.

For energy executives and grid planners, the South Australia experience offers concrete operational data on how battery systems perform under real market conditions: revenue stacking across multiple grid services, degradation rates over multi-year operation, and the regulatory design choices that either enable or obstruct peaker plant displacement.

Key Concepts

Several technical and market design concepts are essential to understanding how the Hornsdale Power Reserve operates and why its results matter for peaker replacement strategies.

Frequency control ancillary services (FCAS): The NEM operates eight separate FCAS markets that compensate generators and storage systems for maintaining grid frequency within the 49.85 to 50.15 Hz band. Battery systems can respond to frequency deviations in less than 200 milliseconds, compared to 5 to 10 minutes for gas turbines, making them far more effective at providing contingency frequency services. Hornsdale generates approximately 55% of its revenue from FCAS markets.

Revenue stacking: Unlike gas peakers that earn revenue primarily through energy dispatch, utility-scale batteries can simultaneously participate in energy arbitrage (charging during low-price periods and discharging during high-price periods), FCAS provision, network support services, and capacity market payments. This layered revenue model is fundamental to the economics of peaker replacement.

Virtual cap contracts: Neoen sells "virtual cap" financial products backed by the Hornsdale battery, providing large energy consumers with price hedging against wholesale electricity spikes above A$300 per MWh. These contracts replicate the function that gas peaker plants traditionally served in financial hedging markets, demonstrating that batteries can substitute for peakers on both physical and financial dimensions.

Inertia and system strength: Synchronous generators like gas turbines inherently provide rotational inertia that helps stabilize grid frequency. Battery inverters do not provide physical inertia but can be configured with grid-forming capabilities to synthetically replicate inertia response. South Australia became the first jurisdiction in the NEM to mandate minimum system strength requirements, and Hornsdale's 2021 upgrade included grid-forming inverter technology to help address this gap.

What's Working

The Hornsdale Power Reserve and the broader South Australian battery deployment have generated results across multiple performance dimensions that other jurisdictions are actively studying.

Peak Price Suppression Is Measurable

Analysis by the Australian National University's Battery Storage and Grid Integration Program found that Hornsdale's energy arbitrage activity reduced the frequency of extreme price events (above A$5,000 per MWh) in South Australia by 74% between 2017 and 2024, saving consumers an estimated A$210 million in wholesale electricity costs (ANU, 2025). The mechanism is straightforward: when demand rises and gas peakers would traditionally set marginal prices at A$5,000 to A$16,600, the battery discharges at lower offer prices, capping the market clearing price. This price suppression effect compounds as additional battery capacity enters the market. Since the 250 MW / 250 MWh Torrens Island Battery (operated by AGL Energy) came online in late 2023, the combined battery fleet has further reduced average peak-hour wholesale prices by an estimated 18%.

FCAS Market Transformation

Before Hornsdale, FCAS costs in South Australia averaged A$47 million annually, with gas generators and interconnector flows providing most frequency control. By 2025, FCAS costs in the state had fallen to A$12 million annually, a 74% reduction. Hornsdale alone provides approximately 35% of South Australia's regulation FCAS requirements. The speed advantage is decisive: the battery responds to frequency deviations within 100 to 200 milliseconds, while gas turbines require 6 to 10 seconds for governor response. The Australian Energy Market Operator has reported that system frequency quality in South Australia improved by 40% following Hornsdale's commissioning, as measured by the time spent outside the normal operating frequency band (AEMO, 2025).

Battery Degradation Is Below Projections

Tesla's original warranty for the Hornsdale Megapack system guaranteed 80% capacity retention after 15 years of operation. Independent monitoring by the University of Adelaide's School of Electrical and Electronic Engineering measured actual capacity retention of 93.2% after five years of high-utilization cycling, exceeding the warranty curve by a significant margin. The system averages 1.5 to 2 full charge-discharge cycles per day, with a cumulative throughput exceeding 380 GWh by the end of 2025. This degradation performance has directly influenced financing assumptions for subsequent battery projects across the Asia-Pacific region, with lenders now applying less conservative degradation curves that improve project economics by 8 to 12% on a net present value basis.

Subsequent Projects Validated the Model

The success of Hornsdale catalyzed a wave of investment. The Victorian Big Battery (300 MW / 450 MWh, operated by Neoen) commenced operations in early 2024. AGL Energy's Torrens Island Battery (250 MW / 250 MWh) replaced a retired gas peaker at the same site, physically occupying a fraction of the former plant's footprint. The Waratah Super Battery (850 MW / 1,680 MWh) in New South Wales, operated by Akaysha Energy, began staged commissioning in 2025 and is the largest grid-scale battery in the Southern Hemisphere. Across the NEM, total installed battery capacity reached 4.2 GW by early 2026, up from 0.2 GW in 2020.

What's Not Working

Despite the overall success, several persistent challenges limit the pace and scale of peaker replacement across the region.

Duration Limitations Constrain Full Peaker Substitution

Most utility-scale batteries in the NEM are configured for 1 to 2 hours of duration at rated power. Peaker plants, while infrequently dispatched, can sustain output for 4 to 12 hours during extended heatwave-driven demand events. South Australia experienced a multi-day heat event in January 2024 where peak demand exceeded 3,200 MW for three consecutive days. Battery systems discharged fully within the first 2 hours of each evening peak, requiring gas peakers to cover the remaining 3 to 4 hours of elevated demand. The Reliability Panel's 2024 review concluded that 4-hour-duration batteries would be needed to fully displace gas peakers for 95% of peak events, but current lithium-ion economics make 4-hour systems 40 to 60% more expensive per MW of capacity than 2-hour systems.

Capacity Market Design Remains Incomplete

The NEM does not operate a formal capacity market. Instead, generators receive revenue only when they dispatch energy or provide FCAS. This energy-only market design creates revenue volatility for battery operators: in years with mild weather and low peak demand, arbitrage revenue can fall by 30 to 50%. The Australian Energy Market Commission has been evaluating a capacity investment scheme since 2023, but implementation has been delayed by jurisdictional disagreements between state governments. Without a capacity payment mechanism, battery projects rely heavily on contracted revenue from virtual cap products and FCAS, which may not provide sufficient long-term certainty to finance the next wave of longer-duration installations.

Transmission Constraints Limit Optimal Siting

Several proposed battery projects in New South Wales and Queensland have encountered connection delays of 18 to 36 months due to transmission network congestion and lengthy generator connection processes. AEMO's 2024 Integrated System Plan identified 14 renewable energy zones across the NEM, but the transmission augmentation required to connect these zones to load centers is behind schedule. Battery projects sited to replace retiring gas peakers at existing connection points avoid these delays, as demonstrated by AGL's Torrens Island Battery, but greenfield sites face significant grid access barriers.

Planning Approval Timelines Are Inconsistent

State-level planning approvals for utility-scale batteries vary significantly across the NEM. South Australia's streamlined assessment process enables approvals within 6 to 9 months for projects under 30 MW, but larger installations require full environmental impact assessments that add 12 to 18 months. Victoria and New South Wales have introduced dedicated assessment pathways for battery projects, but community opposition related to fire safety concerns has delayed several projects by 6 to 12 months, particularly following a thermal runaway event at the Victorian Big Battery during commissioning in 2021 that burned for four days before being contained.

Key Players

Established Companies

• Neoen: French independent power producer that developed and operates the Hornsdale Power Reserve and the Victorian Big Battery, with a combined 494 MW of operational battery capacity across Australia.
• AGL Energy: Australia's largest electricity generator, which replaced a retired gas peaker at Torrens Island with a 250 MW / 250 MWh battery system, demonstrating brownfield peaker-to-battery conversion.
• Tesla: Supplies Megapack battery systems for multiple NEM-connected projects including Hornsdale, and provides long-term operations and maintenance services.
• Origin Energy: Operates the Eraring Battery (460 MW) adjacent to Australia's largest coal-fired power station in New South Wales, positioning battery capacity for post-coal grid support.
• TransGrid: Transmission network operator for New South Wales, responsible for connection agreements and network planning for large-scale battery projects.

Startups

• Akaysha Energy (acquired by BlackRock): Developed the 850 MW Waratah Super Battery, the largest grid battery in the Southern Hemisphere, using a project finance model backed by institutional capital.
• Eku Energy: A subsidiary of ENGIE and Mitsui, developing a pipeline of over 2 GW of battery storage projects across Australia, focused on sites adjacent to retiring thermal generators.
• Edify Energy: Australian renewable energy developer with over 500 MW of co-located solar and battery projects, pioneering hybrid renewable-plus-storage configurations in the NEM.

Investors and Funders

• Clean Energy Finance Corporation (CEFC): Australian government green bank that has committed over A$1.2 billion in debt and equity financing for grid-scale battery projects since 2020.
• BlackRock Climate Infrastructure: Acquired Akaysha Energy in 2022 and has deployed over US$700 million into Australian battery storage through the platform.
• Macquarie Asset Management: Active investor in Australian battery storage through the Macquarie Green Investment Group, with equity positions in multiple NEM-connected projects.

KPI Summary

KPI	Baseline (2017)	Current (2025)	Target (2030)
Hornsdale capacity (MW / MWh)	100 / 129	194 / 259	194 / 259
NEM total battery capacity (GW)	0.1	4.2	12.0
FCAS costs in South Australia (A$ million/year)	47	12	8
Extreme price events (>A$5,000/MWh) frequency reduction	0%	74%	90%
Battery capacity retention after 5 years	N/A	93.2%	92%
Gas peaker dispatches displaced (South Australia)	0%	35%	60%
Average peak-hour wholesale price reduction	0%	18%	30%

Action Checklist

• Conduct a peaker plant utilization audit to identify assets running fewer than 500 hours annually, as these are the strongest candidates for battery replacement based on economics and emissions reduction potential
• Engage with the local transmission or distribution network operator at least 24 months before planned battery commissioning to secure connection agreements and identify any required network augmentation
• Model revenue stacking scenarios across energy arbitrage, FCAS, network support, and capacity payments to assess project viability under multiple market conditions rather than relying on a single revenue stream
• Evaluate brownfield sites at retiring thermal generation facilities, where existing grid connections and site infrastructure can reduce development timelines by 12 to 18 months compared to greenfield locations
• Assess battery duration requirements based on historical peak demand profiles, targeting 2-hour systems for frequency regulation and 4-hour systems for full peaker substitution during extended demand events
• Engage with state and federal financing bodies such as the CEFC or equivalent institutions to access concessional debt that reduces the weighted average cost of capital for battery projects
• Develop community engagement plans that address fire safety concerns proactively, including site design features such as blast walls, fire suppression systems, and setback distances that exceed regulatory minimums

FAQ

Q: How does a battery system actually replace a peaker plant in grid operations? A: Battery systems replace peaker plants by providing the same services: injecting power during periods of high demand and withdrawing when demand falls. The operational difference is response speed and efficiency. A gas peaker requires 5 to 15 minutes to ramp from cold start to full output and operates at 25 to 35% thermal efficiency, meaning 65 to 75% of the fuel's energy is wasted as heat. A battery system can discharge from zero to full rated power in less than one second with round-trip efficiency of 85 to 92%. In capacity markets or energy-only markets, batteries submit offer prices into the wholesale market just as gas peakers do. When demand rises and the market operator dispatches generation in merit order, batteries with lower marginal costs are dispatched ahead of gas peakers, effectively displacing them. The Hornsdale Power Reserve bids into the NEM at marginal costs near zero (since it has no fuel cost), while gas peakers bid at A$80 to A$300 per MWh depending on gas prices.

Q: What is the expected lifespan and replacement cost of a utility-scale battery system? A: Current lithium-ion battery systems used in grid-scale applications carry manufacturer warranties of 15 to 20 years, with guaranteed capacity retention of 70 to 80% at end of warranty. Hornsdale's actual performance data suggests real-world lifespans may exceed warranty assumptions, with capacity retention of 93.2% after five years of intensive cycling. At end of life, battery modules can be refurbished, repurposed for lower-demand applications, or recycled. Replacement costs are projected to decline further: BloombergNEF forecasts lithium-ion pack prices reaching US$80 per kWh by 2030, meaning that mid-life augmentation, adding new modules to restore original capacity, becomes increasingly economical. Several NEM battery projects have included contractual provisions for mid-life augmentation at pre-agreed pricing, protecting against degradation risk while maintaining revenue capacity.

Q: Can this model work in markets that already have formal capacity mechanisms? A: Yes, and in some ways it works even better. Markets with formal capacity mechanisms, such as Japan's capacity market (launched in 2024) or South Korea's capacity payment system, provide batteries with a guaranteed revenue floor for being available during peak periods, reducing the reliance on volatile arbitrage and FCAS revenues. Japan's first capacity auction cleared at approximately JPY 5,000 per kW per year, which alone covers 15 to 20% of the annualized cost of a 2-hour battery system. When combined with energy and ancillary service revenues, batteries in capacity markets can achieve returns on equity of 10 to 14%, compared to 8 to 12% in energy-only markets like the NEM. The key design consideration is ensuring that capacity market rules do not discriminate against storage: some early capacity market designs required participants to demonstrate continuous availability for 8 or more hours, effectively excluding 2- to 4-hour batteries. Japan and South Korea have both revised their rules to allow duration-weighted capacity credits, enabling shorter-duration batteries to participate proportionally.

Q: What role does demand response play alongside battery storage in replacing peakers? A: Demand response and battery storage are complementary rather than competing solutions. In South Australia, the SA Power Networks demand response program enrolled over 85,000 residential smart thermostats and hot water systems by 2025, providing approximately 200 MW of load reduction during peak events. This demand-side capacity extends the effective duration of battery systems: if batteries can cover the first 2 hours of a peak event and demand response reduces the remaining peak by 200 MW, the combined effect can displace a 4-hour gas peaker without requiring 4-hour-duration batteries. AEMO's 2025 Electricity Statement of Opportunities models battery storage and demand response as a combined resource, projecting that the two together can replace 3.8 GW of thermal peaking capacity across the NEM by 2032.

Sources

• Australian Energy Market Operator. (2025). Quarterly Energy Dynamics Q4 2025: Battery Storage Performance and Market Impact. Melbourne, VIC: AEMO.
• Australian Energy Regulator. (2024). State of the Energy Market 2024. Melbourne, VIC: AER.
• BloombergNEF. (2025). Lithium-Ion Battery Pack Prices: 2024 Survey Results and 2030 Outlook. London: BNEF.
• Australian National University Battery Storage and Grid Integration Program. (2025). Impact of Hornsdale Power Reserve on South Australian Wholesale Electricity Market Outcomes: 2017-2024. Canberra, ACT: ANU.
• Australian Energy Market Commission. (2024). Reliability Panel Annual Market Performance Review 2024. Sydney, NSW: AEMC.
• Neoen. (2025). Hornsdale Power Reserve: Operational Performance Report 2020-2025. Paris: Neoen SA.
• Clean Energy Finance Corporation. (2025). Grid-Scale Battery Storage Investment Portfolio: Outcomes and Impact Report. Sydney, NSW: CEFC.
• University of Adelaide School of Electrical and Electronic Engineering. (2025). Long-Term Degradation Analysis of Hornsdale Power Reserve Battery Modules. Adelaide, SA: University of Adelaide.