​In 2025, the average battery energy storage system in ERCOT earned revenues that were 84% lower than 2023's all-time highs.

​This has happened largely for two reasons.

First, 2024 and 2025 saw two consecutive years of relatively mild weather, which contributed to a reduction in extreme price events.

Second, there is likely already enough battery energy storage in ERCOT to manage system reliability, and to keep price spikes at a minimum - for now.

Ancillary services are saturated, while increased battery participation in energy arbitrage is reducing volatility.

The near-term business case has weakened.

However, load growth is coming, and in the meantime, existing thermal generators face the same revenue pressures as batteries. If retirements accelerate before new demand materializes, scarcity will return. And as new demand does materialize, more new generation and storage will be necessary to meet it.

The question isn't whether upside for battery energy storage in ERCOT exists over a 15-20 year horizon - it's how they bridge the gap to that upside in the next 2-4 years.

Key takeaways:

• ​BESS revenues have collapsed due to milder weather and market saturation. BESS capacity grew 70x since 2020, while 2024-2025 lacked the extreme weather that drives scarcity.
• Load growth is real, but overstated. ERCOT's 220 GW headline won't materialize. A realistic projection: 105 GW by 2030, still 22% growth in four years.
• Demand is changing shape. Energy consumption is up 27% since 2019, even as peak demand stagnated. FWEST load grew 116% in six years, 4.3x faster than the grid average.
• Thermal retirements can restore volatility. Over 22 GW of aging coal and gas is at risk. Only combined-cycle gas remained profitable in 2024-2025.
• Bridging solutions are coming. Firming contracts, new ancillary services like DRRS, and creative offtake structures could help projects survive until scarcity returns.

​Less-than-extreme weather and market saturation have reduced battery revenues

2024 and 2025 lacked the extreme weather events relative to seasonal norms that drive scarcity conditions.

The summers in each year were well aligned with the 15-year average from 2008-2022 in terms of average temperature. In other words, temperatures were 'mild' - they didn't deviate strongly from the norm.

​'Mild' doesn't mean below-average, it means a lack of conditions that drive elevated prices.

However, mild weather years aren't the only thing that have suppressed revenue opportunity and capture for BESS in ERCOT in the last two years.

Market saturation is the other driver.

​In fact, there is likely already enough battery energy storage to serve today's challenges in ERCOT - at least the challenges that batteries are equipped to serve, and the market is designed to compensate.

Installed BESS capacity has grown rapidly in ERCOT. Since the beginning of the decade, capacity has grown from ~200 MW to nearly 14,000 MW, representing more than 70x growth. Nearly 10 of those 14 GW came online in 2024 and 2025. Read more about the details of BESS buildout in ERCOT in 2025 and Modo Energy's projections for the future here.

As more batteries connect to the grid, competition increases, driving the cannibalization of revenue opportunities.

​Saturation is real - more battery energy storage increases competition for a relatively stable set and size of opportunities, in both Ancillary Services and Energy.

Read more here to learn more about how the increased presence of BESS has played into declining revenues, and how battery operators have responded.

However, recent revenue opportunities remaining this low is conditional on the current configuration of supply and demand remaining as it is today in the future.

​Load growth and thermal retirements have the potential to reintroduce volatility

​Volatility has nearly evaporated in ERCOT in the last two years. In 2025, there were only three days on which the average battery earned at least $0.50/kW, compared with 16 in 2024 and 58 in 2023.

​However, for volatility - and battery revenue opportunities - to remain this low in the future, the current configuration of supply and demand and recent weather patterns would need to persist. This is not going to happen.

​Load growth is coming, but the timing (and the magnitude) is uncertain

​The headline load growth projections in the ERCOT Long Term Load Forecast are overblown. The ~220 GW peak demand figure cited for 2030 will not materialize. However, meaningful growth is still coming.

The numbers are inflated because the barrier to entry for requesting interconnection as a large load in ERCOT is nearly zero. A prospective data center developer pays nothing to state their intention to develop. Consequently, the queue contains far more prospective demand than will ever materialize, mirroring the pattern seen in generation interconnection.

A more realistic picture emerges by combining a bottom-up appraisal of individual large load projects with a haircut aligned to historical generation throughput rates of approximately 25%.

​This approach projects peak demand reaching approximately 105 GW by 2030. That represents 19 GW of growth over the all time peak of 85.9 GW, or 22%, in just four years.

Several changes to ERCOT's planning guides are moving through the stakeholder process to enable the integration of new types of demand. Planning Guide Revision Request 115, PGRR 134, and other active processes will improve visibility into large load interconnection timelines and may establish higher barriers to entry, potentially requiring deposits of hundreds of thousands of dollars simply to enter the queue.

While the stakeholder process works to create a more realistic picture, it would be a mistake to conclude demand growth is not happening. The headline queue numbers overstate near-term growth, but underlying energy demand is already growing at a meaningful pace.

​How much is demand already growing, and how has the shape of demand changed?

​Peak demand growth stagnated in 2024 and 2025. From 2019 to 2023, peak demand grew from 74,820 MW to 85,508 MW. However, in 2024 and 2025, peak demand declined 1.8%, falling to 83,707 MW.

​Total energy consumption tells a different story. Between 2019 and 2025, total energy grew from 384 TWh to 488 TWh, a 4.08% CAGR. Even as peak demand fell 1.8% between 2024 and 2025, total energy consumption rose 5.8%.

This divergence signals a shift in load profile. Mild weather suppressed peak demand while baseline consumption continued expanding. The growth comes from data centers, residential increases, and continued electrification of the oil and gas industry in the Permian Basin.

New demand - like data centers and electrified O&G - is largely 24x7. This is most visible in the Far West (FWEST) weather zone. FWEST load grew 116.5% between 2019 and 2025, 4.3 times faster than ERCOT overall. The zone now represents 9.2% of total ERCOT load, up from 5.5% in 2019.

​FWEST's load profile is distinctly flat, with a peak-to-trough ratio of just 1.07x compared to ERCOT's 1.34x system average. This flatness reflects continuous data center and O&G operations.

As flat demand combines with storage muting traditional sunset peaks, price spread opportunities shift into later evening hours.

​Some new demand may arrive with on-site gas turbines, but turbine procurement constraints limit this. Not enough turbines are available to absorb all growth.

​Low prices could create a push-pull effect as thermal generators retire

​Low prices and reduced volatility diminish battery revenues, but they’re also detrimental to aging coal and gas generators. These resources often have long minimum run times and startup times. They struggle to justify continuing operating expenses when prices remain low.

Older generators have higher heat rates, requiring more fuel per MWh generated. Their spark spreads, the difference between the electricity price and the cost of fuel needed to generate that electricity, are thinner as a result.

When around-the-clock prices fall, these units face a compounding problem: they cannot cycle quickly to capture price spikes, and they cannot operate profitably during low-price hours.

​As average prices compress, the older, less efficient units are first to become uneconomic, resulting in them running less frequently.

​‘Breakeven’ analysis shows older coal resources are often running at a loss

​'Breakeven' prices can be calculated using full operating costs: fuel costs (combined with assumed heat rates), variable O&M, and fixed O&M converted to $/MWh based on capacity factors.

Older units face higher heat rates and maintenance costs. Over 10 GW of coal generators in ERCOT are more than 40 years old, and 12 GW of gas generators are more than 50 years old.

​In 2024, combined cycle gas generators were the only thermal generators to remain profitable relative to full operating costs. The trend continued in 2025. Higher natural gas prices contributed to ATC prices rising from $27/MWh to $33/MWh. However, older coal's assumed breakeven of ~$36-37/MWh remained above the annual average.

​Retirements could restore volatility

​The eventual retirement of some thermal generation will be partially offset by further wind, solar, and storage additions in the meantime. However, removing dispatchable capacity means prices will become more volatile during periods of low renewable output, unless flexible resource deployment keeps pace.

This creates a push-pull effect: suppressed revenues accelerate retirements, thinning the supply stack, making scarcity more likely, and eventually restoring the volatility that suppressed revenues in the first place.

​Bridging the short-term gap to long-term upside

​Volatility will re-emerge in ERCOT at some point. The question is how projects survive until it does. Three categories of solutions could help bridge the gap: firming requirements, new ancillary services, and creative offtake structures.

​Firming requirements remain undefined

​House Bill 1500, passed during the 2023 Texas legislative session, mandated firming requirements for generation resources in ERCOT.

The requirements apply to generators who sign an interconnection agreement after January 1, 2027, and are applied only to resources that have operated for at least a year.

Generators must operate at or above their average generation capability during high-risk events. Those unable to guarantee meeting performance standards must secure dispatchable capacity through bilateral agreements or co-located buildout.

The PUCT released a proposal in July 2024 for public feedback. Several items remain under discussion.

The proposed Seasonal Average Generation Capability (SAGC) methodology calculates a single average percentage for each hour per season. Stakeholders argue this misrepresents solar output and penalizes thermal generators due to temperature variations.

More critically for BESS, the current proposal makes it difficult for batteries to qualify as firming providers. Only output above SAGC counts toward firming, but battery output typically equals seasonal average. Stakeholders oppose this exclusion, arguing battery flexibility supports grid reliability.

The penalty structure is also contested. The proposal sets penalties at 20% of effective value of lost load (VOLL), capped to 15 critical hours per season. Stakeholders advocate for a fixed $1,000/MWh penalty instead for long-term investment certainty. This would effectively cap potential revenue from a firming contract at $15/kW.

If batteries are included as eligible firming providers, this would create a new bilateral contract revenue source and incentivize co-location with renewables. Projects coming online around 2027 would benefit most.

​New ancillary services could offer diversification

​The only new ancillary service definitively being developed in ERCOT is the Dispatchable Reliability Reserve Service (DRRS). DRRS functions as firm capacity compensation, offering a longer deployment window than Non-Spinning Reserve and deploying ahead of real-time.

However, eligibility will likely be limited to batteries with four or more hours of duration. No batteries active in ERCOT today would qualify.

Other potential services are more speculative. Voltage support service could offer locational value in West Texas by compensating grid-forming inverters for addressing inverter-based resource integration issues. Inertia support service could compensate resources for supporting rate of change of frequency (ROCOF), though eligibility would likely extend to thermal generators.

There is no guarantee these services will be developed. If they are, they would likely be contracted bilaterally rather than through a market. ERCOT could take cues from Germany, where an inertia market is being developed.

​Creative offtake structures

​Traditional tolling agreements have seen limited uptake. Developer needs and offtaker willingness diverge, creating a wide bid-ask spread. Offtakers assign low value to near-term revenue and prefer shorter terms, while developers need longer-term certainty to cover their cost basis.

Several alternative structures could bridge this gap: revenue sharing arrangements with upside participation, virtual or partial tolls, hub-versus-node settling hedges, and capture rate contingent payments.

New offtake partners could also emerge. Insurance underwriters could underwrite revenue against a baseline, whether a flat floor or a moving target pegged to average revenue capture of a market segment.

​Owning, operating, and investing in BESS in ERCOT requires patience - and the right revenue bridge

​The 15-20 year investment case remains viable. Near-term returns are not guaranteed, but the structural conditions for scarcity are likely to re-emerge.

Projects that can bridge the gap through solutions such as:

• maximizing their capture of price spreads,
• the capture of attractive price spreads that extend later into the evening and overnight hours,
• the introduction of new ancillary services,
• entering into contracts that satisfy renewable firming requirements,
• or hedging their exposure to low volatility years through offtake or positions in the forward market,

will be positioned to capture upside when it materializes.