Global Energy Storage: How to power AI with solar-plus-storage The industry has long viewed solar-plus-storage as insufficient for baseload powerparticularly for 24/7 AI-driven demand. Masdar and EWEC’s gigascale solar-plus-storageproject challenges this view with the world’s first deployment of firm renewable power atscale. Our analysis leaves us incrementally more constructive on solar-plus-storage as acompetitive source of baseload power, with Sungrow and CATL as key beneficiaries. Brian Ho, CFA+852 2123 2615brian.ho@bernsteinsg.com Neil Beveridge, Ph.D.+852 2123 2648neil.beveridge@bernsteinsg.com Solar-plus-storage can deliver baseload reliability at scale.UAE’s Masdar and EWECare building 5.2 GW of solar combined with 19 GWh of storage (19 hours duration) todeliver ~1 GW of continuous power, with completion targeted by 2027. Based on ouranalysis, the system could achieve ~99.6% system reliability, representing a structural shiftin renewables from intermittent energy sources to providers of firm capacity. Kelvin Yuan, Ph.D., CFA+852 2123 2612kelvin.yuan@bernsteinsg.com Economics are competitiveness particularly for higher gas price markets.While theupfront capex is high at ~$6,000/kW, solar-plus-storage benefits from low operating costsand zero fuel exposure. At an estimated LCOE of $97/MWh for the project, solar-plus-storage can compete with gas-fired power at gas price of ~$8/mmbtu or higher. At 12-hourstorage scenario, the LCOE falls to around $80/MWh which can still achieve high systemreliability of 95%. The attractiveness of solar-plus-storage as baseload is important givenongoing volatility and disruption in global gas supply. That said, we still believe gas-firedpower is favored in regions with abundant low cost gas supply such as the US. Faster deployment versus gas and nuclear is a key advantage.Solar and storageprojects can be delivered in roughly two years, compared with current gas turbine leadtimes of around four years due to supply constraints, and even longer timelines for nuclear,typically six years or more. The key constraint is access to high solar irradiance and land.While the system isreplicable, we think it is still limited to solar-rich regions with abundant, low-cost land. Thisproject alone requires ~60 km² of land—roughly the size of Manhattan—highlighting thesignificant physical footprint required to deliver firm renewable power at scale. Storage is the primary driver of system economics rather than solar.ESS accountfor roughly half of total project capex, meaning cost competitiveness is primarily driven bystorage cost, efficiency, and performance rather than module pricing. We expect global ESS demand to grow at ~34% CAGR over the next five years,driven by the need for firming renewable generation and grid stability.Within thevalue chain,CATLleads in ESS battery supply and technology, whileSungrowis a keyplayer in system integration, inverters, and broader power conversion solutions, positioningboth as primary beneficiaries of this structural shift. BERNSTEIN TICKER TABLE INVESTMENT IMPLICATIONS In a structurally power-constrained world, solar-plus-storage is emerging as an increasingly important solution to deliver stable,baseload-like power, particularly for fast-growing AI and data center demand. The Masdar project demonstrates that, withsufficient overbuild and long-duration storage, renewables could achieve high system reliability while offering competitiveeconomics in higher gas price environments, alongside the key advantages of being emissions-free and having no fuel pricerisk. While scalability remains dependent on geography, land availability, and grid infrastructure, the overall direction is clearlypositive, with costs expected to decline further through scale and ongoing technology improvements, including next-generationbatteries such as sodium-ion. As storage now drives system economics and demand continues to accelerate, we see strongupside for global ESS deployment and recommend Sungrow and CATL as key beneficiaries of this structural shift. DETAILS The UAE’s Masdar and EWEC project demonstrates at scale that solar-plus-storage can deliver firm baseload power. Thisfundamentally reframes the role of renewables: the key question is no longer whether solar can generate cheap energy, butwhether it can reliably meet a continuous load profile when paired with sufficient storage and overbuild. The project combines5.2 GW of solar with 19 GWh of storage to deliver ~1 GW of continuous power, with completion targeted by 2027. Thisrepresents the first gigascale attempt to overcome solar intermittency through system design, and therefore serves as a keyproof point for the industry. We find that solar-plus-storage is increasingly cost competitive in higher gas price environments.At current cost levels, it can compete with gas-fired generation when fuel prices are elevated, making it an attractive alternativeamid volatility in global gas markets. However, gas rem