Jiangsu Azure Corporation confirmed on 2022-08-23 that its annual cylindrical lithium-ion cell capacity had reached about 700 million units, with the Suzhou-based maker explicitly naming the 4680-format high-capacity cell inside its product portfolio [S1]. On 2022-08-28, Chuneng New Energy broke ground on its Yichang Lithium Battery Industrial Park in Hubei Province, a 60-billion-yuan project that plans to add 150 GWh of lithium battery output on a multi-phase build-out [S5]. These two reference points — one cell-maker scale-up and one industrial-park scale-out — set the operating envelope that process engineers and procurement teams are now modelling against the April 2026 Chinese lithium hydroxide capacity data published by CBCIE [S6].
Capacity planning for lithium covers three distinct layers that must be reconciled before any capital decision: upstream lithium hydroxide output (CBCIE tracks this monthly with YoY and MoM deltas) [S6], mid-stream cell gigafactory ramp (Azure's 700 million cells/yr [S1] and Chuneng's 150 GWh target [S5]), and downstream pack/pack-assembly tooling that follows cell availability. Process engineers typically start with Resource Requirements Planning (RRP), which generates a capacity plan by critical work center on a 12-month-to-3-year horizon, before running Rough-Cut Capacity Planning (RCCP) to identify bottlenecks and Capacity Requirements Planning (CRP) to match personnel and equipment to MRP-generated loads [S3].
Cell-Level Capacity: Cylindrical Format Scale-Up
Azure's 700 million cells/yr, with a stated primary market in power-tool and vacuum-cleaner cylindrical packs and exploration of e-scooter and portable energy-storage applications, gives a concrete benchmark for what a single Chinese cell line delivers at mature utilization [S1]. The 4680-format cell, which Azure lists inside its portfolio, runs at the larger end of the cylindrical family and typically requires higher-format winding and formation equipment than legacy 18650/21700 lines, a fact that drives incremental capex when a plant converts or expands [S1]. For sites where output already targets power tools, the cell format mix often clusters in 18650 and 21700 diameters; planners should confirm the exact format-share before sizing formation, aging, and DCIR-test capacity.
Overseas customers account for more than 50% of Azure's lithium-battery revenue, which means a Chinese cell-capacity build must be paired with shipping, UN 38.3 transport certification, and IEC 62133-2 cell-level safety documentation before the line can be loaded with export orders [S1]. The 4680 line in particular forces a re-check of formation cycler count, since a single 4680 cell carries roughly 4-5× the Ah of a 21700 cell, and a fixed-MWh formation floor needs proportionally fewer pieces — a sizing fact that planners frequently miss when they scale by cell-count rather than by MWh throughput.
Park-Level Capacity: 150 GWh Industrial Build-Outs
The Chuneng New Energy Yichang Lithium Battery Industrial Park plans to build 150 GWh of lithium battery production capacity in Longquan Town, Yiling District, Yichang City, with the project officially starting construction on 2022-08-28 [S5]. At 60 billion yuan total investment across the full phased build, the implied capex is roughly 0.4 yuan/Wh of nameplate — a benchmark number that procurement teams can use to sanity-check vendor quotations on similar greenfield Chinese gigawatt-hour projects. The 150 GWh figure is nameplate at full ramp, not commissioning throughput, so planners should pair it with a phased-loading curve rather than treating it as Year-1 output.
Industrial-park scale at 100 GWh+ shifts the bottleneck away from the cell line itself and toward anode/cathode precursor supply, electrolyte tank farm (where the dosing flow meter on each transfer line sets the achievable concentration tolerance), and formation DC power. A 150 GWh park typically requires a dedicated lithium hydroxide line upstream — which is precisely the segment that CBCIE tracks monthly for April 2026 with YoY(%) and MoM(%) indicators [S6] — and a single-source LiOH dependency is the most common failure mode when greenfield parks commission behind schedule.
Upstream Feed: Lithium Hydroxide Capacity as the Constraint
Production Capacity of Lithium Hydroxide in China, reported monthly by CBCIE, gives the upstream throughput envelope that any cell-capacity plan must consume before it can ramp [S6].
Lithium hydroxide is preferred over lithium carbonate for high-nickel cathode chemistries (NMC811, NCA) because hydroxide routes avoid the high-temperature re-crystallization step that carbonate feeds require. The geopolitical dimension is also material: lithium supply is more geographically dispersed than fossil-feedstock supply, which the academic literature frames as a structural change in energy geopolitics and a partial driver of capacity-planning complexity in the Li-ion chain [S4].
Planning Method: RRP → RCCP → CRP Stack
The standard three-tier capacity-planning stack — Resource Requirements Planning (RRP), Rough-Cut Capacity Planning (RCCP), and Capacity Requirements Planning (CRP) — is the right scaffolding for any lithium gigafactory model [S3]. RRP runs against a long-term forecast and answers facility-expansion, new-site acquisition, staffing-load, and capex questions on a 12-month-to-3-year horizon; RCCP then flags constraints at critical work centers; CRP finally matches available personnel and equipment to MRP-generated load and indicates whether the plan must be revised or resources added [S3]. For a Chinese cell plant, the critical work center is almost always formation/DCIR-test or, on older lines, electrolyte filling under dry-room dewpoint control — those are the RCCP flags that drive a CRP loop.
The JD Edwards-style flowchart logic is platform-agnostic: a lithium park planner should always generate the resource-requirements plan after the long-term forecast and before the Master Scheduling program runs, because RRP is the input that validates the monetary allotment in the strategic business plan [S3]. Skipping RCCP and going directly from RRP to a purchase order is the most common root cause of stranded capex in greenfield cell lines.
Decision Criteria: Matching Capacity Layer to Investment Horizon
A capacity decision at the cell, park, and upstream layer carries different time horizons, capex intensity, and reversibility, and the comparison below is the table most procurement and EPC teams need before signing. [S1]
<strong>Cell line (700 M cells/yr reference):</strong> capex per Wh moderate, ramp 9-18 months, reversible by re-tooling format, key risk is overseas-customer mix and 4680-format formation cycler count [S1].
<strong>Industrial park (150 GWh reference):</strong> capex roughly 0.4 yuan/Wh nameplate, ramp 3-5 years phased, partially reversible, key risk is upstream LiOH single-source dependency and anode/cathode precursor supply [S5].
<strong>Lithium hydroxide (CBCIE national figure):</strong> capex per ton high, ramp 24-36 months, sunk cost, key risk is MoM/YoY swing in the monthly national capacity series and high-nickel cathode-chemistry fit [S6].
Failure Modes and Constraints
The dominant failure mode in 2026 lithium capacity planning is a non-reconciled upstream-midstream model: a 150 GWh park announcement is paired with a LiOH supply contract that is shorter or smaller than the park's nameplate, and the gap surfaces only at commissioning [S5][S6]. A second failure mode is format mis-spec: planners scale by cell-count rather than MWh and end up with a formation floor that is undersized for 4680 cells, which carry roughly 4-5× the Ah of a 21700 cell [S1]. A third failure mode is geopolitical: lithium resource geography is more dispersed than legacy fossil-feedstock geography, and the academic literature explicitly notes that this dispersion shifts the planning problem from resource concentration to logistics and trading-hub design [S4].
Process engineers also need to watch dry-room dewpoint for electrolyte filling — with the pressure sensor on the dry-room recirculation line typically setting the trip threshold — formation DC-power demand at cold-start, and UN 38.3 plus IEC 62133-2 cell-level certification for export flows above 50% overseas-customer mix [S1]. A useful cross-reference for industrial plant instrumentation tied to capacity planning is the CPU Process Control and Instrumentation 2026 spec band guide, which lines up PLC and PCB architecture choices against the same scale-up envelope a lithium park faces. The upstream feedstock side of the bill of materials is covered in Rare Earth Key Components in 2026 Bill of Materials, which sits one step further back from LiOH but influences the same precursor-throughput assumption. For a broader material-availability lens that frames lithium alongside other critical inputs, the Rare Earth Raw Material Sourcing Guide 2026 is a useful comparator.
Sourcing Signals to Track Through End-2026
Two trackable signals will resolve most of the residual uncertainty in a 2026 lithium-capacity plan. Second, Azure's next investor-platform update on the 4680-format ramp will confirm whether the 700-million-cell base is being re-allocated toward 4680 or held in 18650/21700 mix, and the answer will set the formation-cycler demand signal for the second half of 2026 [S1]. A third, lower-frequency signal is the next Chuneng New Energy Yichang phase-completion announcement, which will convert part of the 150 GWh nameplate into commissioned MWh and tighten the upstream LiOH demand pull [S5].