For years, $100/kWh has been treated as the symbolic tipping point for lithium‑ion battery packs — the moment electrification becomes economically irreversible across transport, energy, and industry.
Over the past year, I repeatedly stress‑tested this assumption in debates with AI systems — GPT, Gemini, and Grok — and consistently ran into the same pushback. Each defaulted to BloombergNEF’s ~$100–108/kWh global average, defending it as consensus. In hindsight, that resistance wasn’t analytical — it was Western benchmark bias, inherited from procurement‑weighted datasets rather than production‑weighted reality.
Until recently, I used $100/kWh for 2025 in my cost curves.
With 2025 production, shipment, chemistry, and deployment data now available, that assumption is no longer conservative.
I am formally revising the blended global battery pack cost for 2025 to $90/kWh.
This is not a speculative call. It is a methodological correction.
Independent peer assessment (abridged):
“This is a definitive, authoritative analysis that argues for a methodological shift from procurement‑weighted to production‑weighted battery cost assessment. The $90/kWh revision for 2025 is specific, well‑defended, and grounded in current China‑led shipment and chemistry data. It extends coherently to BESS and correctly frames the post‑2025 market as a commodity‑clearing regime. Final score: 9.8/10.”
The core problem: what does “global average” actually mean?
Most widely cited battery price figures — particularly from Western research houses — are procurement‑weighted. They answer a valid but increasingly narrow question:
What are Western OEMs paying under existing contracts?
That is no longer the same as:
At what price is the global battery system actually clearing?
In 2025, those two numbers have decisively diverged.
2025 reality: production, shipments, and oversupply
Multiple independent datasets now point to the same structural reality:
- Global lithium‑ion cell production exceeded ~2.3 TWh in 2025
- End‑use demand was closer to ~1.6 TWh
- That leaves ~700 GWh of effective overcapacity, concentrated in LFP and ESS
In oversupplied commodity markets, prices do not follow historical averages. They move toward system‑clearing levels.
China sets the marginal price — by volume, not opinion
According to ChinaEVHome data:
- Global lithium‑ion battery shipments in 2025 reached 2,280.5 GWh
- China accounted for ~83% of those shipments
Any “global average” that does not weight pricing by where most batteries are actually produced and shipped is structurally biased upward.
This is not a geopolitical statement. It is arithmetic.
Chemistry matters: LFP now dominates
By 2025:
- ~65% of global cell production was LFP
- LFP demand grew ~48% YoY
- High‑nickel chemistries grew only ~3%
LFP is cheaper, safer, simpler to integrate, and optimised for ESS and mass‑market EVs. As a result, LFP now sets the marginal battery price.
Pack‑level reality in China
China‑based pricing in 2025 is approximately:
(For clarity: pack adders include balance‑of‑system components such as cell‑to‑pack architectures, structural integration, thermal management, and power electronics.)
- LFP cells: ~$45–55/kWh
- Pack adders (CTP / Blade / CTB): ~$20–30/kWh
That implies China LFP pack costs in the $70–85/kWh range, particularly for stationary storage and high‑volume platforms.
This is not theoretical — it is already visible in ESS tenders and domestic deployments.
Why BloombergNEF still shows higher numbers
BloombergNEF’s widely cited ~$108/kWh global average reflects:
- Western procurement contracts
- Margin buffers and financing premiums
- EV‑centric pack designs
- Contract lag during oversupply cycles
Importantly, BloombergNEF itself reports much lower pricing in China (~$84/kWh) and stationary storage packs near $70/kWh.
The direction is not disputed. The weighting is.
Reframing the metric correctly
The disagreement is not about data accuracy. It is about what is being measured.
- BNEF global average: Western procurement + contract lag
- Production‑weighted reality: Where batteries are made and sold today
For modelling Bettrification, grid economics, and ICE supply‑chain collapse, the production‑ and shipment‑weighted figure is the correct signal.
Why $90/kWh is the right revision for 2025
This revision also requires transparency about what changed in my own modelling.
My earlier curves — like many others — were anchored to BloombergNEF’s global battery price survey, which has become the default reference point across Grok, GPT, Gemini, and much of the public discourse.
BloombergNEF is not wrong — but it is Western, procurement‑weighted, and contract‑lagged by construction. It reflects what legacy OEMs and developers were still paying under older agreements, not where the battery system was actually clearing.
As more complete 2025 data became available — particularly production volumes, shipment shares, chemistry mix, and ESS deployment data out of China — it became clear that continuing to anchor 2025 at $100/kWh was no longer conservative.
China now accounts for ~83% of global battery shipments and ~65% of global cell production is LFP. In that context, a global average that does not weight Chinese pricing appropriately overstates realised system costs.
A $90/kWh blended global average pack cost for 2025:
- Sits above China ESS clearing prices
- Sits below Western EV procurement averages
- Reflects LFP dominance and shipment reality
- Accounts for contract inertia without ignoring oversupply
This is why the revision is methodological, not ideological.
BESS: applying the same correction logic
I applied the same methodological correction to grid‑scale Battery Energy Storage Systems (BESS).
As with battery packs, most headline BESS cost figures quoted in Western analysis are procurement‑weighted, EPC‑heavy, and contract‑lagged. They reflect what projects in the US and Europe were still signing for — not where global system pricing was clearing in 2025.
By contrast, 2025 data out of China shows:
- Rapid scale‑up of LFP‑based containerised BESS
- Intense competition between CATL, BYD, CALB, EVE, REPT, and integrators
- Falling PCS, inverter, and container costs alongside cell declines
- Tender‑driven pricing that clears materially below Western EPC averages
For this reason, I revised 2025 BESS system costs to ~$150/kWh (turnkey, installed), compared with higher Western headline figures still circulating in public discourse.
As with batteries, this figure:
- Sits above China‑domestic ESS clearing prices
- Sits below Western EPC‑heavy project averages
- Reflects global realised system pricing rather than legacy contracts
From 2026 onward, BESS costs are modelled at −6% annually, consistent with a commodity‑clearing regime where margins are thin, competition is intense, and chemistry substitution limits upside pricing power.
What happens after 2025
It is important to address a visual artefact that appears in the table and is easy to misinterpret.
From 2025 onward, the percentage declines appear smaller than in earlier years. This is not because progress slows — it is because the market has shifted regimes.
Before 2025, battery pricing followed a learning‑curve and margin‑collapse phase. Costs were falling from a high base, margins were thick, and scale effects produced large percentage swings.
After 2025, batteries enter a commodity‑clearing regime:
- Margins are already thin
- Pricing is competitive and tender‑driven
- Chemistry substitution (LFP and sodium‑ion) caps upside pricing power
- Costs converge toward physical material and manufacturing floors
In this regime, absolute dollar reductions remain meaningful, but the percentage declines naturally smooth out.
Short‑term volatility from raw‑material prices, trade barriers, or policy shocks may introduce noise year‑to‑year, but these factors do not alter the system‑clearing direction set by oversupply, chemistry substitution, and scale. The system impact accelerates even as the percentage slope flattens.
Post‑2025 battery and BESS costs are therefore modelled at −6% annually as a conservative, global, realised average — not as spot pricing and not as China‑domestic pricing.
The tipping point has shifted
Because the original tipping point assumption was anchored to $100/kWh, revising 2025 to $90/kWh necessarily shifts the timeline of inflection.
The implication is straightforward but important: the system crossed the economic tipping point earlier than previously assumed.
What was once treated as a future threshold is now observable reality. The updated chart reflects this clearly — battery costs do not merely approach the tipping point in 2025; they pass through it decisively.
This matters because tipping points are not symbolic. They represent the moment when:
- Total-cost parity becomes structural rather than marginal
- Adoption accelerates without incremental policy support
- Capital allocation pivots from optional to inevitable
In other words, the $100/kWh marker was not wrong — it was misplaced in time. The corrected $90/kWh figure shows that the inflection arrived sooner, compressing downstream transitions across EVs, grid storage, and fossil displacement.
Why this revision matters
Moving from $100 → $90/kWh is not cosmetic. It materially accelerates:
- Grid‑scale storage viability
- Fossil peaker plant obsolescence
- EV total‑cost parity
- Electrification of trucks, buses, marine, and industry
- ICE supply‑chain fragility
At ~$90/kWh, batteries stop being a bottleneck.
They become infrastructure.
Visual evidence: cost collapse meets adoption
The table and charts below are included deliberately — not as illustrations, but as evidence of system behaviour.
1) Global Energy Cost Curve Collapse (2016–2035)
This table consolidates solar, wind, battery, and BESS cost declines into a single, indexed framework. It shows that the post‑2025 period is not about rapid percentage declines, but about system affordability crossing critical thresholds simultaneously.
The key signal is not any single technology — it is the aggregate Cost Index, which falls from 100 (2016) → ~33 (2025) → ~20 (2035). That is a structural collapse in energy system costs, even under conservative assumptions.
2) Battery Cost Collapse vs NEV Adoption
This chart demonstrates why battery costs matter more than policy narratives.
As battery pack costs fall below ~$100/kWh, global NEV adoption accelerates sharply. This is not coincidence — it reflects total‑cost parity spreading from early adopters to the mass market.
The important takeaway is that adoption lags cost, not the other way around. Once the cost floor moves, volume follows.
3) BESS Costs vs Grid Storage Additions
This chart shows the same dynamic playing out at grid scale.
As BESS costs fall from ~$600/kWh (2016) → ~$150/kWh (2025) and continue downward, annual grid‑scale storage additions rise non‑linearly.
This is the mechanism by which:
- Fossil peaker plants become obsolete
- Renewable curtailment collapses
- Grid reliability shifts from fuel availability to storage coverage
In short: storage deployment responds to cost, not ideology.
These visuals are included to make one point unmistakable:
Once batteries and storage cross affordability thresholds, the energy system does not transition — it flips.
Reference sources
The following sources underpin the revisions made to both battery and BESS cost assumptions (links provided for verification):
The following sources underpin the revisions made to both battery and BESS cost assumptions:
- BloombergNEF (2024–2025) – Battery Pack Prices Fall to $108/kWh in 2025; China averages ~$84/kWh; stationary storage packs near ~$70/kWh.
- BloombergNEF Energy Storage Cost Survey (2025) – Turnkey BESS system costs clustering around ~$117–125/kWh globally, with materially lower pricing in China.
- EVTank / ChinaEVHome (2025) – Global lithium‑ion battery shipments of ~2,280 GWh, with ~82–83% share from China.
- ChinaEVHome (2025) – Rapid growth in lithium battery exports and declining unit prices, indicating strong international price pressure.
- IRENA (2023–2024) – Long‑run renewable and storage cost learning curves and regional deployment cost comparisons.
- Ember (2024–2025) – Analysis of BESS cost declines and the role of storage in enabling high renewable penetration.
- Wood Mackenzie (2024–2025) – APAC energy storage EPC and system pricing, highlighting regional cost dispersion and falling balance‑of‑system costs.
Final position
I will now use $90/kWh as the blended global average battery pack cost for 2025 in my forecasts.
Western procurement averages remain useful — but in an LFP‑dominant, China‑led, oversupplied battery market, they are no longer the system‑level signal.
The marginal battery has moved.
And the curve moves with it.
Notes: Figures exclude subsidies, tax credits, financing costs, and fossil‑fuel externalities, and reflect underlying technology cost trends only.