Lithium carbonate futures touched CNY 194,000 per tonne on May 8, up 197% year-on-year, as BESS and AI data centre demand displace EVs as the marginal price driver. AMG's Bitterfeld hydroxide refinery and Moment Energy's Texas second-life gigafactory, announced within days of each other, reveal how the industry is reorganising around the structural supply gap that price signal implies. Together, the three developments describe a battery economy that is simultaneously tighter, more geographically distributed, and more strategically contested than at any point since the 2022 price peak.
Introduction
On May 8, 2026, Chinese lithium carbonate futures on the Guangzhou Futures Exchange settled at CNY 193,983 per tonne, a figure that requires some historical framing to appreciate fully. At the trough of the preceding correction cycle, on June 23, 2025, the same contract traded at CNY 58,400 per tonne. The move from that floor to the current level represents a 232% appreciation in under eleven months, and the year-on-year comparison of roughly 197% places the current price among the most acute lithium rallies on record outside the 2022 peak.
UBS, in a May 7 research note, raised its 2026 average price forecast by 18% to CNY 200,000 per tonne inclusive of VAT, while flagging a scenario in which spot prices reach CNY 250,000 in May or June. The bank's base-case demand estimate of 1.97 million tonnes of lithium carbonate equivalent for 2026 implies 16% year-on-year growth, but its risk-weighted supply forecast of 1.91 million tonnes LCE produces a structural gap of roughly 65,000 tonnes. Morgan Stanley's parallel deficit estimate sits at approximately 80,000 tonnes LCE.
Three developments announced in the first week of May crystallise exactly how that gap is being experienced across different parts of the value chain. AMG Critical Materials reported Q1 2026 results on May 6 showing its Bitterfeld battery-grade lithium hydroxide refinery in Germany producing in-specification product and generating approximately EUR 20 million in quarterly revenue from a standing start. On May 5, British Columbia-based Moment Energy announced a $40 million Series B to build a 200,000-square-foot second-life battery gigafactory in Taylor, Texas. And the lithium carbonate price itself, reaching CNY 194,000 on May 8, validated both corporate strategies simultaneously.
Building on the analysis of BESS and AI infrastructure demand published in this column in May 2026, the convergence of these three data points is not coincidental. They are three expressions of the same structural realignment: surging BESS demand driven by AI infrastructure build-out, a supply side constrained by geopolitical resource nationalism and permitting delays, and an industry responding by diversifying both the geography of refining and the definition of supply itself.
The Price Signal: BESS and AI Replace EVs as the Marginal Demand Driver
The composition of the current lithium rally is as important as its magnitude. In prior cycles, EV adoption provided the dominant demand narrative. The 2026 price structure is materially different. Global EV sales in Q1 2026 declined approximately 3% year-on-year, with weakness in China partially offset by European momentum. That a 3% volume contraction in the largest end-use segment coincides with a 197% year-on-year price increase is the clearest possible indication that demand leadership has passed to another buyer category.
That category is BESS, increasingly driven by AI data centre operators. Benchmark Mineral Intelligence data shows BESS demand grew 51% in 2025, against 26% growth in EV-related battery demand, with BESS reaching 19% of total Chinese battery demand by year-end and hitting 45% in December alone as year-end installation activity peaked. Benchmark projects new operational BESS capacity exceeding 450 GWh in 2026, against 315 GWh installed in 2025. UBS's upside scenario assumes 80% year-on-year ESS demand growth, which would create a deficit of 123,000 tonnes LCE at an implied price of CNY 250,000 per tonne.
The AI data centre dimension is structural rather than cyclical. As Moment Energy's announcement noted, power availability has become the primary constraint on data centre expansion across North America, and BESS installations co-located with hyperscale facilities require more lithium per unit of output than passenger EV batteries. Trading Economics and Fastmarkets both cite fresh bulk buying from data centre operators as a distinct price driver in the current move, separate from EV or grid-utility procurement channels.
The policy overlay reinforces the demand arithmetic. Beijing's commitment to doubling national EV charging capacity to 180 GW by 2027 will itself require substantial BESS deployment to manage grid load variability. BYD's upward revision of its 2026 overseas sales forecast to 1.5 million units, from 1.3 million in January, adds incremental cathode demand. And UBS notes that Chinese demand for electric heavy trucks is expanding rapidly, a segment with higher battery content per vehicle than passenger cars. The three demand vectors are not competing; they are additive.
The Supply Constraint: Jianxiawo, Zimbabwe, and the Shrinking Offer Window
The demand acceleration is meeting a supply side that is structurally less elastic than the headline reserve numbers would suggest. Two specific disruptions account for a meaningful share of the current tightness, and their interaction with longer-run refining geography creates a compounding effect.
CATL's Jianxiawo lepidolite mine in Yichun, Jiangxi, which produces approximately 65,000 tonnes of lithium carbonate equivalent annually and accounts for roughly 3% of global supply, suspended operations on August 9, 2025, following the expiration of its mining permit. While a post-Chinese New Year restart was anticipated, sources have increasingly pointed to a materially later resumption, potentially in the second half of 2026. Benchmark's analysis indicates that a sustained Jianxiawo delay, layered onto existing Jiangxi regional disruptions, would push the global market from a 78,000-tonne LCE surplus to a deficit, warranting price reassessment across 2026 and H1 2027.
The Zimbabwe situation, which this column examined in detail in the April and May 2026 pieces on spodumene transit disruption and the sovereign turn in critical minerals, adds a second constraint of comparable magnitude. Zimbabwe's February 25 export suspension covers approximately 7% of global 2026 LCE supply, per Fastmarkets, with Chinese customs data showing that Zimbabwe contributed 1.204 million tonnes, or roughly 15% of total Chinese spodumene import volumes, in 2025. BMI has revised its 2026 Zimbabwe mine production forecast to 131,100 tonnes LCE, forecasting tightened supply through at least mid-to-late 2027 pending commissioning of local processing capacity.
The market is pricing these constraints in real time. The validity window for battery-cell price offers has shrunk from three months to as little as 14 days, according to Fastmarkets, as producers decline to carry price risk across a period in which spot carbonate can move 9% intraday, as it did on February 26 immediately following the Zimbabwe announcement. Andy Leyland of SC Insights summarises the arithmetic: 2026 demand growth at 24%, supply growth at 19%. The gap has a price.
AMG Bitterfeld: The Mid-Stream Refining Premium in a Deficit Market
AMG Critical Materials' Q1 2026 results, released May 6, provide the clearest corporate-level data point yet on the advantage accruing to vertically integrated producers with operational mid-stream refining in the current environment. The lithium segment reported revenue of $60.6 million, up 89% year-on-year, as the Bitterfeld battery-grade lithium hydroxide plant generated approximately EUR 20 million in quarterly revenue from a facility that was not contributing commercial output at the same point in 2025. Average realised spodumene concentrate prices reached $916 per tonne CIF China in Q1, up 43% year-on-year, while production costs fell from $572 per tonne to $417, compressing unit costs as volumes rose 11%.
The Bitterfeld plant's significance extends beyond AMG's income statement. It represents one of the very few battery-grade lithium hydroxide refining assets operating outside China's dominant processing complex. China controls an estimated 80% of global lithium-ion battery production and a virtual monopoly on conversion capacity. European automakers and cell producers seeking to comply with emerging supply chain provenance requirements for subsidy eligibility face a structurally limited set of non-Chinese hydroxide suppliers. AMG's planned expansion to 130,000 tonnes per year of total capacity by end-2026, processing Brazilian spodumene into German-refined product, directly addresses that geographic gap.
The company is simultaneously extending the circularity of its refining model. Engineering work has begun on a 5,000-tonne lithium carbonate-to-hydroxide conversion plant at Bitterfeld designed to accept recycled lithium carbonate feedstock, at a capital cost of approximately $50 million, 20% of which is supported by a German federal grant announced in December 2025. The project underscores a point that the broader market is beginning to price: in a supply-constrained environment, the ability to accept diverse and recycled feedstocks is a competitive differentiator, not merely a sustainability credential.
Management reiterated full-year 2026 adjusted EBITDA guidance of $210 to $240 million and noted that Q2 EBITDA should approach Q2 2025 levels, supported by peak tantalum prices and favourable shipment phasing in the lithium segment. With spot spodumene at approximately $2,450 per tonne and lithium hydroxide at $20,600 per tonne as of late April, the earnings trajectory for H2 2026, as Bitterfeld reaches full commercial loading, is considerably more positive than the guidance language implies.
Moment Energy and the Second-Life Supply Layer
The most conceptually significant of the three May developments is also the smallest in absolute capital terms. Moment Energy's $40 million Series B, bringing total funding to over $100 million, is not material relative to the multi-billion-dollar capital cycles of conventional gigafactory construction. What it represents is the first credible, certified, at-scale commercial deployment of a business model that partially bypasses the upstream lithium supply chain entirely.
The company's model is straightforward in principle and technically complex in execution. EV battery packs retaining 70% to 80% of original capacity after 10 to 20 years of automotive use are disassembled, graded, tested, and reassembled into containerised BESS units rated for up to 30 years of further operation, at cycling costs as low as 3 cents per kWh for industrial users. The Taylor, Texas gigafactory, a 200,000-square-foot facility already broken ground upon, is planned to produce up to 1 GWh of storage annually, the largest second-life battery manufacturing operation in the United States. The workforce target of 250 reflects a capital-light model: Moment Energy's management notes that building new lithium-ion battery factories at competitive scale requires several billion dollars, while the Taylor facility costs a fraction of that by leveraging chemistry that was already produced and paid for.
The investor syndicate is instructive about the demand thesis. The Series B was led by Evok Innovations and included Liberty Mutual Investments, W23 Global Fund, Acario (the corporate venture arm of Tokyo Gas), Amazon's Climate Pledge Fund, Voyager Ventures, and In-Q-Tel. The presence of an insurance company, a Japanese utility, and a defence-adjacent technology fund alongside Amazon's climate vehicle reflects the breadth of the customer base: hyperscale data centres, industrial operators, airports, and utilities, all seeking BESS supply that is FEOC-compliant, domestically manufactured, and insurable under standard policies. CEO Edward Chiang's statement that the company is building energy infrastructure that can be deployed rapidly and manufactured domestically captures the supply chain security dimension that makes second-life batteries a strategic procurement option rather than merely a cost-optimisation exercise.
The market context matters here. CATL and its Chinese peers currently account for approximately 72% of global BESS supply, according to BNEF. For US buyers navigating Foreign Entity of Concern compliance requirements under the Inflation Reduction Act and related Treasury guidance, the pool of eligible BESS suppliers is materially smaller than the pool of technically available systems. Moment Energy's UL 1974 and UL 9540A certifications, which management emphasises make it the only provider able to deploy second-life BESS in the built environment without special dispensations, are not incidental to the funding story. They are the monetisable regulatory moat.
Strategic Implications: Refining Geography, Circular Feedstocks, and the Emerging Supply Stack
The three May data points, read together, describe an industry that is constructing a new supply architecture in real time. The first layer of that architecture is the price signal itself: CNY 194,000 per tonne carbonate, UBS forecasting CNY 200,000 as a 2026 average with CNY 250,000 as a plausible peak, and a deficit consensus ranging from 22,000 to 123,000 tonnes LCE depending on ESS demand assumptions. That range of outcomes alone is sufficiently wide to require strategic positioning across multiple scenarios.
The second layer is mid-stream refining geography. AMG Bitterfeld's EUR 20 million quarterly revenue contribution from a newly operational plant validates the thesis, articulated in this column's May analysis of CATL's multi-chemistry announcements, that the value in the current cycle is migrating from concentrate to refined product. Producers who can deliver battery-grade hydroxide with non-Chinese provenance documentation are capturing a premium that the spot carbonate price alone does not fully reflect. The planned expansion to 130,000 tonnes per year, combined with the recycled carbonate conversion unit, positions Bitterfeld as a dual-feedstock refinery capable of serving both primary and circular supply chains, a configuration that is likely to become more competitively relevant as EU battery regulation provenance requirements tighten.
The third layer is the second-life supply contribution. Moment Energy's 1 GWh annual Taylor output is small relative to the 450-plus GWh of new BESS capacity Benchmark projects for 2026, but the trajectory matters more than the current volume. CEO Chiang's stated mission to repurpose all retired EV batteries by 2030 is an aspirational target, but the underlying feedstock arithmetic is compelling: North American EV penetration is now sufficiently advanced that the first generation of mass-market EV battery packs is beginning to exit their automotive service lives at scale. Each pack diverted to second-life BESS represents lithium, cobalt, and nickel demand that never reaches the upstream spot market.
AMG's own recycled carbonate-to-hydroxide conversion project at Bitterfeld reflects the same logic applied to European feedstock flows. Together, the two circular models suggest an emerging supply category that operates orthogonally to the mining cycle, neither correlated with spodumene prices nor exposed to export bans and permit delays. Whether that category scales quickly enough to materially affect the deficit projections for 2026 and 2027 is uncertain; the volume is too small for that effect in the near term. The strategic signal is that the largest capital allocators in the BESS supply chain, from Amazon and Tokyo Gas to the German federal government, are funding that optionality deliberately.
Forward Outlook
The UBS base case of CNY 200,000 per tonne as a 2026 average implies further upside from the current CNY 194,000 level before year-end mean reversion. The bank's upside scenario of CNY 250,000, contingent on 80% ESS demand growth, does not require extraordinary assumptions: 51% BESS demand growth was achieved in 2025, and the AI infrastructure build-out accelerating across North America and Europe represents a demand increment that was not embedded in most 2024 forecasts. The Jianxiawo restart timing and Zimbabwe's conditional quota system, which as of April 2026 offers partial but structurally inadequate relief, are the primary variables that determine where within the UBS range the market settles.
For producers, the near-term priority is navigating the offer validity compression. A 14-day price validity window is operationally disruptive for cathode makers accustomed to quarterly contracts, and at least one has already turned to the spot market for feedstock as a direct consequence of Zimbabwe-related pipeline gaps. Producers with locked-in long-term supply agreements and in-house refining, AMG being the clearest western example, are insulated from that volatility in a way that pure-concentrate sellers are not.
For downstream buyers, the FEOC compliance calendar is tightening. The combination of Chinese BESS market dominance at 72% of global supply, rising lithium prices driven substantially by Chinese BESS demand itself, and US regulatory incentives favouring domestically manufactured storage creates a structurally awkward procurement environment. Second-life solutions from certified domestic producers represent one exit from that constraint. Mid-stream refining from non-Chinese provenance suppliers represents another. Neither is yet sufficient at the scale the market requires, but both are receiving the capital needed to grow.
The structural picture heading into H2 2026 is one of a market that has absorbed Zimbabwe's supply shock, is carrying forward Jianxiawo uncertainty, and is simultaneously processing BESS demand growth that outpaced even bullish 2025 forecasts. The price at CNY 194,000 per tonne is not speculative; it is the clearing level for a market that has run out of convenient buffer. The three May data points, AMG's Bitterfeld revenues, Moment Energy's Texas groundbreak, and the futures curve itself, are all pointing in the same direction: the second structural lithium cycle is not approaching. It has arrived.