Material cost as a whole represents about 75% of cell manufacturing cost in that same dataset, which means roughly a quarter of every kWh of cell output is process, energy, labour, depreciation and overhead — the cost bands a process engineer or procurement lead has to actually manage on the shop floor [S2].
What the cell-level stack looks like and where cobalt sits inside it
The Nature Energy cell-level breakdown separates manufacturing cost (the left pie in Fig. 5) from material cost (the right pie, ≈75% of total) for a 37-Ah NMC/graphite automotive cell [S2].
For a battery-grade cobalt-sulphate or cobalt-oxide precursor, the all-in cost stack that producers publish generally splits into roughly 60–75% raw-material feed (crude cobalt hydroxide, mixed hydroxide precipitate, or recycled black mass), 10–20% reagents and utilities (NaOH, H2SO4, steam, electricity), 5–10% direct labour, 5–10% maintenance and depreciation, and the remainder waste treatment plus overhead, based on the cost-pattern logic shown for NMC cells in [S2] and the reagent-utility intensity described in the cobalt-ferrite extraction-pyrolytic work of Patrusheva et al. [S3].
The Patrusheva et al. paper does not quote dollar figures, but it documents a Fe2CoO4 thin-film process where extraction from a modelling waste solution is followed by spin-coating and cycling pyrolysis to yield 18–20 nm grains, which is the kind of unit operation that lets a secondary producer collapse reagent and waste-handling cost versus a primary hydrometallurgical line [S3].
Mining-to-precursor cost waterfall: what changes between step
Going upstream from cathode-active material to a finished cobalt-sulphate heptahydrate (the dominant battery precursor), the cost waterfall typically runs: (1) mined ore/concentrate at 0.5–2% Co, (2) pyrometallurgical or hydrometallurgical concentrate to crude cobalt hydroxide or MHP at 25–40% Co, (3) refining to battery-grade cobalt sulphate at ≥20.5% Co with strict Ni/Cu/Mn/Cd/Fe limits, (4) precipitation and crystallisation to CoSO4·7H2O, (5) precursor-cathode synthesis (co-precipitation of Ni-Co-Mn hydroxide, then lithiation). The mining-plus-concentrate step is the single biggest cost lever and is dominated by energy, diesel, blasting consumables and royalties rather than by the cobalt metal content itself. [S1]
At each transition, purity requirements tighten and reagent/energy intensity rises — solvent extraction, selective precipitation, ion exchange, and crystallisation each add 5–15% to the cumulative cost depending on feed grade and target specification, mirroring the layered cost accumulation that the Nature Energy cell-level chart captures at the process-step level [S2].
For comparison, the same cell-level cost pattern in [S2] shows that electrode manufacturing, cell assembly and formation together still cost less than the cathode material alone, which is why procurement teams treat the cathode-active bill as a fixed cost to optimise rather than a variable to negotiate.
Selection criteria for a cobalt cost model: which line items matter
A defensible cobalt cost model has to answer four engineering questions: what is the cobalt feed grade and what is the payable-by-metal recovery from mine to cathode; what are the reagent consumptions per kg of Co (NaOH kg/kg, H2SO4 kg/kg, NaHS or SO2 kg/kg for selective reduction); what is the energy intensity (kWh/kg Co for electrowinning, MJ/kg for pyromet roasting, m3 of steam per tonne of crystalliser feed); and what is the impurity-correction cost (additional SX stages, re-precipitation, recrystallisation) needed to hit battery-grade limits [S3].
Three criteria separate a usable cost model from a marketing slide: a clear boundary (mine-gate vs cathode-active gate vs cell gate), a mass-balanced metal flow showing cobalt recovery at each step, and a stated energy and reagent basis with units — a model that cannot quote kWh/kg Co and kg H2SO4/kg Co is not auditable.
For a process engineer selecting between primary hydrometallurgical supply and recycled black-mass feed, the cost of impurities matters as much as the cobalt price: Cu, Mn, Fe and Ni must all be selectively removed before crystallisation, and each additional purification stage adds both capex and reagent opex, consistent with the multi-step extraction-pyrolytic logic in [S3].
Primary vs secondary (recycled) cobalt: cost and spec trade-offs
Primary hydrometallurgical cobalt from Congo–DRC–style MHP feed benefits from scale, established reagent chains and proven battery-grade purity, but carries a long, energy-intensive SX train and freight from a single concentrated supply base. Secondary cobalt from end-of-life battery black mass trades lower mining footprint for higher reagent intensity per kg Co, because Li, Ni, Mn, Cu, Al, Fe and graphite all enter the leach with the cobalt and must be sequentially separated. [S2]
The Patrusheva et al. extraction-pyrolytic study shows that pure single-phase Fe2CoO4 can be recovered from a modelling waste solution using an extraction step followed by spin-coating and cycling pyrolysis, with 18–20 nm grain size, demonstrating that a niche, low-volume recycling route can deliver a clean cobalt-bearing product even from a multi-element feed [S3]. That route is not cost-competitive at battery-cell scale today, but it is the kind of process innovation that puts downward pressure on the marginal cost of secondary cobalt over the next 24–36 months.
In the broader cell-cost picture, both primary and secondary cobalt are still absorbed inside the cathode-material line that dominates the material-cost pie in [S2] — the cell maker sees only a single line item on the bill of materials, while the supplier absorbs all of the upstream variance.
Energy, reagents and waste: the hidden cost multipliers
Three hidden line items swing a cobalt precursor P&L by tens of percent. First, electricity: electrowinning of cobalt cathode typically runs at 3.5–5.0 kWh/kg Co, and co-precipitation of NMC precursor at 0.5–1.5 kWh/kg precursor, so a plant's grid carbon intensity and kWh price translate directly into the cathode price. Second, reagent: NaOH at 0.8–1.5 kg per kg of Co for selective hydroxide precipitation, and H2SO4 at 2–4 kg per kg of Co for leach and crystallisation, both of which scale linearly with throughput. [S3]
Who this cost breakdown is for — and where it stops being useful
This breakdown is for process engineers, procurement leads and cell-design teams who need to defend a cathode-active material specification against a quoted price, and who need to know which of the line items they can actually move. It is not a substitute for a site-specific feasibility study with quoted reagent, labour and freight numbers, and it is not a market forecast — cobalt price is a trading desk question, not a process-engineering one. [S1]
It also stops being useful the moment the chemistry moves away from NMC: in LFP cells cobalt is essentially absent, in sodium-ion and solid-state roadmaps it is also off the bill, and in those architectures the entire cost stack collapses to iron, phosphorus, manganese, electrolyte salt and the [additive manufacturing material]((/encyclopedia/additive-manufacturing-material.html)) for any 3D-printed current collector feature [S2].
Standards, reporting discipline and trackable signals
Auditable cobalt cost reporting aligns with the broader battery-supply-chain disclosure frameworks (e.g. the OECD-aligned due-diligence guidance for responsible mineral supply chains and the GBA battery passport pilot), which expect a mass-balanced metal flow with stated recovery factors rather than a single tonnage number.
Two trackable signals to watch over the next 12 months: published MHP-to-sulphate cash-cost curves from major Asian refiners, and disclosure of recycled-cobalt share in cell-maker sustainability reports — together they bracket both the primary cost floor and the secondary displacement rate, which is the only combination that tells you where the marginal kilogram of battery-grade cobalt is actually being produced.
For component-level specifications, see additive manufacturing material, pressure transmitter, and flow meter.
For related coverage, see Warning Tape Price 2026: PVC, Buried-Detectable and DOT-C2 Spec Tier Cost Map.