Batteries are now cheaper than natural gas turbines used at many data centers

Four-hour battery storage was cheaper than open-cycle gas turbines in all 43 markets surveyed by Wood Mackenzie, which expects battery electricity costs to fall while gas-turbine costs rise. AI data-center demand has driven up turbine prices and procurement times: open-cycle units now take two to four years to procure, while closed-cycle turbine waitlists extend into the early 2030s. The report projects four-hour batteries will be 33% cheaper by 2035 in the Middle East and Africa; U.S. solar faces tariff and import-restriction pressure, with 168 GW largely protected by tax-credit safe-harbor provisions.

Analysis

The key investment implication is a change in the marginal build choice, not an imminent collapse in gas demand. Lower-cost four-hour storage can undercut new open-cycle peakers where peak needs are short and charging power is available; it does not replace round-the-clock generation, extended-duration backup, or gas plants already operating. Turbine scarcity may therefore support near-term pricing and value for existing dispatchable capacity even as it weakens the economics of some future peaker projects.

The report’s cost comparison needs a system-level check: verify whether it captures interconnection, charging costs, degradation, replacement, and the value of firm capacity. Those factors can erase an apparent project-level advantage, especially in constrained grids. Falling storage costs also need not translate into battery-maker profits; competition may pass savings through to project owners, while tariffs and supply-chain constraints could interrupt that pass-through in the U.S.

Over the next 1–3 months, look for utility and data-center procurement announcements, project cancellations or deferrals, and evidence that storage is securing interconnection and firm-capacity contracts. Over 6–18 months, sustained battery cost declines could redirect new peaker investment toward storage and renewables, pressuring turbine order expectations. The contrarian risk is over-extrapolating a four-hour cost comparison into a broad gas-demand bear case: data-center load growth and longer-duration reliability needs may support gas even as storage takes peak-share.

Key Decisions for Investors

• Prefer selective exposure to grid-scale storage developers and integrators over a broad clean-energy basket; require evidence of contracted revenue, interconnection access, and project economics before treating lower battery costs as earnings growth.
• Avoid a blanket short of gas producers or utilities. Near term, turbine backlogs and rising peak demand can support incumbent dispatchable assets; the more exposed bucket is prospective open-cycle peaker construction that loses to storage on full project economics.
• Treat U.S. solar as a differentiated watch item: tariff pressure could raise costs for unprotected projects, while safe-harbored projects may retain an execution advantage. Verify project-level eligibility and equipment sourcing rather than extrapolating protection across the sector.
• Falsify the storage-substitution thesis if utility procurements continue to favor gas despite comparable all-in costs, storage projects fail to secure interconnection or capacity payments, or updated analysis shows materially higher costs after charging, degradation, and firming. Reassess gas exposure if turbine orders or data-center power plans weaken.