UBS just upgraded China’s BESS forecast to 666GWh by 2030. Here’s what that means for clean energy, AI data centres, and storage economics worldwide.
China’s BESS Boom: What UBS’ Upgrade Really Means for Clean Power and AI
China is now expected to install 666GWh of battery energy storage systems (BESS) by 2030, according to UBS’ latest forecast. That’s not just a big number on a slide deck. It’s a signal that grid-scale batteries are shifting from “pilot projects” to core infrastructure – especially in a world driven by AI, data centres and 24/7 clean power demand.
Most companies get this wrong. They see storage as an optional add-on to solar and wind. China’s policy shift – and UBS’ upgrade – shows the opposite: standalone BESS is becoming its own asset class, with its own revenue stack, economics and role in the energy system.
This matters because green technology is no longer just about producing clean electrons. It’s about controlling when, where and how those electrons are used. AI, data centres, and electrified industry need reliability, not just “more renewables.” Storage sits right in that gap.
In this post, I’ll break down what’s actually changing in China’s BESS market, why UBS is suddenly more bullish, and what it means if you’re:
- A developer or investor looking at storage deals
- A data centre or industrial offtaker chasing 24/7 green power
- A company building or buying AI and digital infrastructure
1. Policy Shift: From Mandates to Market Signals
China’s recent policy changes are pushing energy storage away from rigid mandates and towards market-driven returns.
The key trigger is Document No. 136, issued by China’s National Development and Reform Commission (NDRC) and National Energy Administration (NEA). It removed the blanket requirement that every new renewable project include a fixed amount of storage.
On paper, that sounds like a headwind for batteries. In reality, it’s the opposite.
What Document 136 really does
Here’s the thing about Document 136: it doesn’t kill storage – it forces it to stand on its own economics.
- No more “you must add X% storage” rules to connect renewables
- Stronger push toward market-based competition instead of subsidies and fixed tariffs
- Opens the door for standalone BESS projects to play directly in power markets
UBS analyst Yishu Yan argues that the BESS industry will actually “ride the policy tailwind” because:
- Storage projects can now stack more diverse revenue streams
- Regulators are widening peak-valley price spreads (day vs night prices)
- Capacity price mechanisms and capacity compensation are being formalised
In plain language: fewer blunt mandates, more explicit price signals that reward storage for doing what it does best – time-shifting and stabilising the grid.
Why this is a big deal for green technology
For the broader green technology space, this shift is healthy:
- It encourages smarter deployment of storage, not just box-ticking for permits
- It promotes efficient grid operation rather than overbuilding renewables
- It creates a template other markets can copy as they open capacity and ancillary markets
If you’re planning projects outside China, watch this closely. This is exactly the kind of market design upgrade every renewables-heavy grid will need.
2. Standalone BESS vs Solar-Plus-Storage: Who Wins?
UBS expects standalone BESS to outgrow solar-plus-storage in China, both in scale and in revenue diversity.
Most early storage in China was tethered to solar or wind farms. That made sense when the main goal was to help renewables meet grid requirements and smooth output. Under the new rules, that’s no longer the only or best play.
Why standalone BESS is pulling ahead
Standalone BESS can tap into revenue streams that are either limited or unavailable to solar-plus-storage hybrids:
- Auxiliary services (frequency regulation, voltage control, reserve)
- Capacity leasing or capacity payments from grid operators
- Spot market arbitrage – buying when power is cheap, selling when it’s expensive
- Capacity compensation mechanisms designed specifically for firm capacity
Solar-plus-BESS can still do arbitrage and smoothing, but:
- It’s often constrained by the solar generation profile
- It can be locked into long-term tariffs that limit upside
- It’s physically and commercially tied to a single generation asset
Standalone systems, by contrast, are grid-centric, not plant-centric. They sit where the grid needs them most and can respond to multiple markets and signals at once.
What this means for project developers and investors
If you’re developing or funding storage projects, this shift suggests a few hard truths:
- You can’t treat BESS as a solar accessory anymore. It needs its own model, risk analysis and optimisation strategy.
- Revenue stacking is non-negotiable. Projects that only rely on one revenue stream will struggle.
- Location and grid context matter more. The best projects will sit at nodes with real congestion, volatility and ancillary service demand.
Developers who still think in “PV-plus-storage packages” only are going to leave value on the table.
3. The New Economics: From 2.5% IRR to 13%+ with Smart Design
The economics of BESS in China are improving fast, driven by both technology costs and smarter market design.
UBS lays out a simple but powerful point: price spreads and capacity payments matter more than most pro formas assume.
How price spreads change BESS returns
UBS models a typical BESS project and shows how changes in the peak–valley price spread reshape returns:
- At a spread of RMB 0.25/kWh (~US$0.035/kWh), the internal rate of return (IRR) sits around 2.5% – barely attractive.
- If that spread widens to RMB 0.4/kWh, IRR climbs to about 8%.
- Add capacity compensation on top, and IRR can reach roughly 13%.
That’s not a rounding error. That’s the difference between:
- Projects that only work on subsidies or balance sheet financing
- Projects that attract infrastructure capital at scale
For regulators and grid operators, the conclusion is blunt: if you want private capital to build storage, you need price structures that reward flexibility.
LCOS, cycle life and why utilisation matters
On the cost side, China’s lithium battery supply chain is doing what it does best: pushing costs down.
- UBS notes that the theoretical cost of solar-plus-BESS in China has reached about RMB 0.3/kWh in 2025, comparable to thermal generation.
- But real-world levelised cost of storage (LCOS) is still higher, around RMB 0.4–0.5/kWh, because utilisation is low.
A stark data point: standalone BESS in China averaged only 248 cycles per year in 2024. That’s less than one full charge–discharge cycle per day.
The reality? Storage systems don’t get cheaper just because batteries do. They get cheaper when they’re used more intelligently:
- Higher utilisation (more cycles per year)
- Better dispatch optimisation with advanced software and AI
- Participation in multiple markets instead of a single use case
This is exactly where AI and green technology intersect:
- AI-powered optimisation can decide when to charge, discharge or provide services minute-by-minute
- Predictive analytics can maximise revenue while protecting battery health
If you’re in energy software or AI, this is a massive opportunity: make those 248 cycles per year look more like 350–400+, and project economics change overnight.
4. AI & Data Centres: The Surprise Storage Super-Driver
China’s AI and data centre boom is turning battery storage from a nice-to-have into core infrastructure for digital growth.
Artificial intelligence isn’t just an IT story. It’s an energy story first. Training large models and running hyperscale data centres demands huge, constant power – and increasingly, that power needs to be verifiably green.
UBS highlights a critical trend: AIDC (artificial intelligence and data centres) are:
- Pushing for direct green power connections to renewables
- Exploring off-grid or near-grid solutions to bypass bottlenecks
- Demanding high reliability and multi-hour backup
What AIDC wants from storage
For these facilities, the bar is high:
- Energy storage allocation ratios of at least 25–30% of their connected renewables
- Storage durations of 4+ hours to ride through variability and peak prices
That’s a very different profile from the 1–2 hour batteries often tied to solar plants.
In practice, this means:
- Bigger batteries, longer duration
- More complex control strategies (grid + behind-the-meter optimisation)
- Tighter integration between IT load, on-site generation and storage
If you’re building AI infrastructure, you can’t treat power as a simple utility bill line item. You’re effectively designing your own micro energy system, and BESS is the backbone.
Why this doubles demand for grid-connected storage
UBS expects that as AIDC applications scale, they could double the demand for grid-connected storage paired with renewables.
That fits the wider pattern we’re seeing in green technology:
- Digital growth (AI, cloud, edge) is fusing with clean energy infrastructure
- Sustainability targets are no longer just ESG slides – they’re tied to power contracts, PPA structures and storage capacity
- Corporates want both green electrons and controllable capacity – and storage is the only way to guarantee both at once
If you’re in the data centre or AI space and you’re not already modelling 4–8 hour storage into your growth plans, you’re behind.
5. UBS’ New Forecast: 666GWh in China, 1,045GWh Globally by 2030
UBS has upgraded both its China and global BESS installation forecasts, signalling a faster, larger build-out of storage than previously expected.
Here are the key numbers:
- China 2025 BESS forecast: upgraded to 150GWh
- China 2026 BESS forecast: upgraded to 232GWh (a 7–19% increase over prior estimates)
- China 2030 BESS forecast: 666GWh, at a 27% CAGR from 2027–2030
Globally, UBS now expects:
- 2025 global BESS installations: 276GWh
- 2026 global BESS installations: 412GWh
- 2030 global BESS installations: 1,045GWh
China’s share of that is enormous. If these numbers hold, China alone would account for well over half a terawatt-hour of installed storage capacity by 2030.
What this means for your strategy
If you’re working in green technology, these forecasts aren’t just trivia; they’re a planning horizon:
- Supply chains: Component makers, integrators and software providers should assume a world where gigawatt-hour projects are standard, not exceptional.
- Skills and talent: The market will need far more people who understand BESS design, optimisation and operations, not just PV or wind.
- Capital allocation: Infrastructure funds and corporates that treat storage as “niche” will struggle to match the returns of those who bake it into their core portfolios.
There’s a better way to approach this: stop thinking of BESS as a project add-on and start treating it as the missing infrastructure layer that makes high-renewables, AI-heavy economies possible.
How to Act on This: Practical Moves for 2026 Planning
If you’re looking at 2026 budgets or strategy right now, here’s how to turn China’s BESS shift into something actionable, wherever you operate:
-
Separate your BESS business case from your generation business case.
Model standalone BESS with its own revenue stack: arbitrage, ancillary services, capacity, and behind-the-meter value. -
Invest in optimisation, not just hardware.
The difference between 248 cycles per year and 350+ is usually software and operations, not cell chemistry. -
For data centres and AI workloads, set explicit storage targets.
Don’t just set “renewables %” goals. Define MWh of storage and minimum duration per MW of load (e.g., 30% / 4 hours) and plan infrastructure accordingly. -
Watch policy, not just prices.
Capacity markets, flexibility tenders and new ancillary services products will decide whether your storage projects earn 3% or 13%. -
Treat storage as part of your climate and resilience strategy.
For corporates, BESS isn’t only about saving on bills; it’s about business continuity, emissions reduction and access to more volatile grids.
The broader story for our Green Technology series is clear: AI, renewables and storage are converging into one system, not three separate industries. The UBS China forecast is just one more proof point.
If this is where China is heading by 2030, the real question is simple: will your energy strategy be ready for a world where storage is everywhere – and expected?