Peaker plant replacement & capacity markets KPIs by sector (with ranges)

Natural gas peaker plants supplied approximately 10% of US electricity generation capacity in 2024 but operated at capacity factors below 10%, running only during periods of extreme demand while contributing disproportionately to grid emissions and consumer costs. Across Asia-Pacific, the economics are shifting even faster: Australia's Hornsdale Power Reserve battery demonstrated that a 150 MW lithium-ion system could replace peaker plant functions at 30% lower cost while responding 4,000 times faster than gas turbines. As utilities, grid operators, and regulators evaluate the transition from fossil-fueled peakers to clean alternatives, the KPIs used to evaluate these replacements determine whether projects deliver genuine reliability improvements or merely shift risk to ratepayers. This article provides sector-specific benchmark ranges drawn from 2023-2025 deployments across Asia-Pacific, North America, and Europe, with guidance on separating meaningful performance metrics from vanity statistics.

Why It Matters

Peaker plants represent one of the most economically vulnerable segments of the fossil fuel fleet. These facilities, typically open-cycle gas turbines or older reciprocating engines, earn revenue primarily through capacity payments rather than energy sales. In the US alone, ratepayers spend approximately $12-15 billion annually on capacity payments to peaker plants, many of which are 30-50 years old and operate at heat rates 40-60% worse than modern combined-cycle units. In Australia's National Electricity Market, peaker gas generation set marginal prices during fewer than 200 hours per year in 2024 but influenced wholesale costs across thousands of additional settlement intervals.

The replacement opportunity is driven by three converging factors. First, battery energy storage costs have declined 89% since 2010, with utility-scale lithium-ion systems now available at $250-350 per kWh installed for four-hour duration, according to BloombergNEF. Second, capacity market reforms in PJM, ISO-NE, AEMO, and other system operators increasingly recognize storage and demand response as eligible capacity resources with performance obligations equivalent to thermal generation. Third, air quality regulations are tightening around aging peaker facilities, many of which are located in environmental justice communities and emit nitrogen oxides (NOx) at rates 5-10 times higher per MWh than baseload gas plants.

For engineers evaluating replacement projects, the KPI framework matters because poorly chosen metrics can mask fundamental performance gaps. A battery system that clears capacity auctions based on nameplate power rating but cannot sustain output for the required duration fails to deliver equivalent reliability. Conversely, over-specifying duration requirements based on historical peaker run times that modern grid operations will never replicate wastes capital on unnecessary storage capacity.

Key Concepts

Capacity Factor vs. Availability Factor. Peaker plants historically operated at capacity factors of 2-10%, meaning they generated electricity only 175-876 hours per year. Their value derived from availability, being ready to generate on short notice during system emergencies. Replacement resources must demonstrate equivalent or superior availability, typically measured as the percentage of hours the resource can deliver rated capacity within contractual response times. Battery storage systems routinely achieve availability factors of 97-99%, compared to 85-92% for aging gas peakers that experience forced outages from mechanical failures.

Effective Load Carrying Capability (ELCC). ELCC quantifies how much firm capacity a resource contributes to system reliability, accounting for its availability, duration, and correlation with peak demand periods. A 100 MW four-hour battery does not provide 100 MW of capacity credit in most market frameworks. ELCC values for four-hour batteries range from 50-95% of nameplate capacity depending on the grid's peak duration profile. Systems with longer, flatter peaks (common in tropical Asia-Pacific climates) assign lower ELCC to short-duration storage than systems with sharp, brief peaks.

Levelized Cost of Capacity (LCOC). Unlike levelized cost of energy (LCOE), which measures the cost per MWh generated, LCOC measures the cost per MW-year of capacity provided. This metric properly captures the economics of peaker replacement because the primary service being procured is reliability, not energy. Comparing LCOC across technologies normalizes for differences in capacity factor, fuel cost volatility, and operational profiles.

Ramp Rate and Frequency Response. Battery storage systems can ramp from zero to full output in milliseconds, compared to 5-15 minutes for simple-cycle gas turbines and 2-5 minutes for aeroderivative turbines. This superior response capability provides additional grid services beyond simple capacity, including primary frequency response and synthetic inertia, which aging peakers cannot deliver.

Peaker Replacement KPIs: Benchmark Ranges by Sector

Utility-Scale Battery Storage (Replacing Gas Peakers)

Metric	Below Average	Average	Above Average	Top Quartile
Installed Cost ($/kWh, 4-hr)	>$400	$300-400	$250-300	<$250
Round-Trip Efficiency	<82%	82-87%	87-91%	>91%
Availability Factor	<95%	95-97%	97-99%	>99%
Degradation Rate (%/yr)	>3%	2-3%	1.5-2%	<1.5%
Capacity Market Revenue ($/kW-yr, PJM)	<$40	$40-70	$70-100	>$100
Response Time (zero to full output)	>1 sec	250ms-1s	100-250ms	<100ms
LCOC ($/kW-yr, 20-yr life)	>$120	$85-120	$60-85	<$60
ELCC (% of nameplate, 4-hr)	<55%	55-75%	75-90%	>90%

Demand Response Programs (Peak Shaving)

Metric	Below Average	Average	Above Average	Top Quartile
Enrolled Capacity Utilization	<60%	60-75%	75-90%	>90%
Customer Participation Rate	<25%	25-45%	45-65%	>65%
Average Curtailment per Event (%)	<10%	10-20%	20-30%	>30%
Program Cost ($/kW-yr)	>$80	$50-80	$30-50	<$30
Dispatch Reliability (% of called events)	<85%	85-92%	92-97%	>97%
Peak Reduction Achieved vs. Target	<70%	70-85%	85-95%	>95%

Virtual Power Plants (Aggregated Distributed Resources)

Metric	Below Average	Average	Above Average	Top Quartile
Aggregation Cost ($/kW enrolled)	>$200	$120-200	$60-120	<$60
Dispatch Accuracy (% of target MW)	<80%	80-90%	90-95%	>95%
Telemetry Latency	>5 sec	2-5 sec	0.5-2 sec	<0.5 sec
Customer Churn Rate (%/yr)	>20%	12-20%	6-12%	<6%
Revenue per Enrolled kW ($/yr)	<$30	$30-60	$60-100	>$100
Capacity Credit (% of enrolled)	<40%	40-60%	60-80%	>80%

What's Working

Australia's Hornsdale and Victorian Big Battery

The Hornsdale Power Reserve in South Australia, originally 100 MW / 129 MWh and expanded to 150 MW / 194 MWh, has become the global reference case for peaker replacement economics. In its first five years of operation, the facility generated approximately AUD $200 million in revenue across energy arbitrage, frequency control ancillary services (FCAS), and network support. The Australian Energy Market Operator (AEMO) documented that Hornsdale's sub-second response capability reduced FCAS costs in South Australia by AUD $116 million in its first two years alone. The Victorian Big Battery (300 MW / 450 MWh), operational since 2023, has demonstrated that the model scales effectively, achieving 98.7% availability and clearing capacity obligations in all called intervals during the 2024 summer peak season.

PJM Capacity Market Integration

PJM Interconnection, the largest US wholesale electricity market serving 65 million customers, cleared 3,200 MW of battery storage in its 2025/2026 capacity auction, up from 450 MW in the 2022/2023 auction. The clearing price of $49.49/MW-day ($18,064/MW-year) for the MAAC region provided sufficient revenue for four-hour battery systems to achieve 8-12% unlevered returns when combined with energy arbitrage and ancillary services. PJM's Capacity Performance construct, which imposes severe penalties for non-delivery during performance assessment intervals, has actually favored batteries over aging gas peakers: battery non-performance penalties totaled $1.2 million across the fleet in 2024, compared to $47 million for gas peakers experiencing forced outages during the same events.

Japan's Grid Stability Programs

Following the 2022 power supply crunch that nearly caused rolling blackouts in the Tokyo Electric Power Company (TEPCO) service territory, Japan's Organization for Cross-regional Coordination of Transmission Operators (OCCTO) accelerated procurement of storage-based capacity resources. By 2025, 4.2 GW of battery storage was operational or under construction across Japan, with Sumitomo Electric's 50 MW / 300 MWh vanadium redox flow battery in Hokkaido demonstrating that long-duration storage can provide capacity services through Japan's extended summer and winter peak periods, which frequently last 6-10 hours.

What's Not Working

Duration Mismatch in Extended Heat Events

Four-hour battery storage performs well for sharp demand peaks but struggles during extended heat events lasting 8-12 hours, increasingly common across South and Southeast Asia. During India's 2024 heat wave, peak demand in the Northern Region exceeded forecasts for 14 consecutive hours on multiple days, a duration that would exhaust any economically viable lithium-ion system. Grid operators in tropical climates must plan for peak events that exceed typical four-hour battery durations, requiring either longer-duration storage technologies (flow batteries, compressed air, thermal storage) or hybrid approaches pairing batteries with dispatchable clean generation.

Interconnection Queue Delays

In both the US and Australia, battery storage projects face interconnection study backlogs that delay commercial operation by 3-5 years. AEMO's 2024 Integrated System Plan identified 28 GW of storage in the connection pipeline but projected that only 8 GW would achieve commercial operation by 2030 due to network hosting capacity constraints and transmission augmentation requirements. In PJM, the average time from interconnection request to commercial operation for battery projects exceeded 48 months in 2024. These delays create a gap between committed peaker retirements and replacement resource availability that poses genuine reliability risks.

Capacity Market Design Flaws

Several capacity market frameworks inadequately value the attributes that make batteries superior to peakers. Markets that clear on nameplate capacity without adjusting for ELCC overvalue short-duration storage. Markets with insufficient performance penalties fail to differentiate between reliable batteries and unreliable aging peakers. The ISO-NE capacity market's treatment of energy storage has been particularly problematic: the minimum offer price rule (MOPR) initially prevented state-supported storage projects from clearing at competitive prices, effectively protecting incumbent gas peakers from competition. While ISO-NE has since reformed MOPR, the episode illustrates how market design can delay economically and environmentally superior alternatives.

Meaningful Metrics vs. Vanity Metrics

Vanity Metric: Nameplate Capacity Added. Press releases announcing MW of battery storage deployed tell you nothing about reliability contribution. A 100 MW one-hour battery provides far less system reliability than a 50 MW four-hour system.

Meaningful Metric: ELCC-Adjusted Capacity. ELCC captures the actual reliability contribution accounting for duration, availability, and correlation with system peak. Engineers should benchmark against ELCC values published by system operators (AEMO, PJM, CAISO) rather than vendor nameplate claims.

Vanity Metric: Capacity Auction Clearing Volume. Clearing a capacity auction indicates market participation, not operational performance. What matters is performance during actual stress events.

Meaningful Metric: Non-Performance Penalty Exposure. The ratio of penalty payments to capacity revenues reveals whether a resource delivers on its reliability commitment. Top-performing battery systems achieve penalty-to-revenue ratios below 1%, while aging peakers frequently exceed 10%.

Vanity Metric: Installed Cost per kWh. Low installed cost is meaningless if the system degrades rapidly, requires expensive augmentation, or achieves low round-trip efficiency.

Meaningful Metric: LCOC over Asset Life. Levelized cost of capacity over a 15-20 year asset life, including degradation, augmentation costs, and residual value, provides the true economic comparison between peaker alternatives.

Action Checklist

• Evaluate local peaker fleet age, heat rates, and emissions profiles to identify highest-priority replacement candidates
• Model ELCC for candidate replacement technologies using system operator methodologies specific to your grid
• Calculate LCOC across a 20-year horizon including degradation, augmentation, and decommissioning costs
• Assess peak demand duration profiles for the past 10 years to determine minimum storage duration requirements
• Engage with capacity market operators early to understand eligibility requirements, performance obligations, and penalty structures
• Evaluate hybrid configurations pairing batteries with demand response or other clean capacity for extended peak events
• Factor interconnection timelines into project schedules, allowing 36-60 months from application to commercial operation
• Monitor capacity market design reforms that may alter the competitive position of storage versus gas peakers

Sources

• BloombergNEF. (2025). Battery Energy Storage System Costs and Market Outlook, Q1 2025. New York: Bloomberg LP.
• Australian Energy Market Operator. (2024). 2024 Integrated System Plan. Melbourne: AEMO.
• PJM Interconnection. (2025). 2025/2026 RPM Base Residual Auction Results Report. Valley Forge, PA: PJM.
• Hornsdale Power Reserve. (2024). Five-Year Operational Performance Summary. Adelaide: Neoen Australia.
• International Energy Agency. (2025). Batteries and Secure Energy Transitions. Paris: IEA Publications.
• Lawrence Berkeley National Laboratory. (2024). Utility-Scale Battery Storage: Empirical Analysis of Project Cost, Performance, and Pricing Trends in the United States. Berkeley, CA: LBNL.
• Organization for Cross-regional Coordination of Transmission Operators. (2025). Japan Electricity Supply-Demand Outlook and Storage Integration Status. Tokyo: OCCTO.
• Lazard. (2025). Lazard's Levelized Cost of Storage Analysis, Version 10.0. New York: Lazard.