The latest modelling for asset allocation in battery storage projects has a new iteration: V8. This develops the idea of a technology transition cliff — a rapid market shift — to embrace the importance of non-disruptive events as well as disruptive ones. Sang Bong Song outlines the thinking.

It was the last item on the agenda, and the room had thinned out. A capex committee at a mid-sized European industrial group — the kind that owns both a small high-nickel cathode pilot line and a portfolio of telecom backup contracts in Africa — was working through its 2027 plan. 

Two proposals sat side by side on the projector.

The first was a roughly $500 million expansion of a high-nickel NMC line tied to a multi-year automotive supply contract, complete with the usual gated qualification milestones. 

The second was a similarly sized program to refresh thousands of lead-acid telecom backup strings across a regulated network, with annual recompetition and a gradient of incremental AGM (absorbent glass mat) and lead-carbon upgrades behind it. 

The finance director put it plainly: “We are running the same Gantt chart on both. Are we sure that’s right?” 

That question — whether the same financial planning architecture applies to a cliff-shaped chemistry transition and a gradient-shaped retrofit — is one a new piece of public-disclosure analysis lets us answer more honestly than before. 

The asymmetry of published findings 

Industry analysis has a publication-bias problem. Reports detail the so-called “cliffs” they find — the point where, the inflection points they identify and the transitions they call — but very rarely publish the chemistries where their model came up empty. 

That is convenient for headlines but corrosive for capital allocation, because the negative result — the regime where no cliff can be identified — is exactly the information a finance committee needs to choose between a step-change playbook and an incremental ladder. 

The independent IDP V8 framework, released on Zenodo on May 21, 2026, is one of the first public-disclosure tools to put negative findings on equal footing with positive ones. 

In one sentence: V8 supplies a verdict on whether a technology trajectory is cliff-shaped, and it reports as a first-class output the trajectories where no cliff can be identified — including memory semiconductors, the Bondalti chlor-alkali membrane series, and several perovskite photovoltaic trajectories. 

That four-tier rubric (S/A/B/C/D), with Tier D explicitly defined as “framework-boundary negative findings,” is the V8 contribution this piece is built around. 

Not every battery transition is a cliff 

V8 finds clean cliff signatures in three battery families. NMC cell-level specific energy is a Tier A cliff with the framework’s discrete-step verdict pinned at the upper bound, consistent with the qualification-gate behaviour the automotive supply chain has been planning around for several years. LFP cell-level specific energy is also a Tier A cliff, with a cliff location estimated around 2021 and a wide confidence band.

Solid-state battery (SSB) trajectories sit in the framework’s “compatible architecture, awaiting mass-production sample” position — cliff-shape strongly suggested, full Tier A activation pending real production data. 

Three further battery regimes that any senior executive in this readership knows well share the structural features V8 associates with Tier D clusters, even though V8 has not yet fit them directly.

For vanadium redox flow specifically, V8’s framework-boundary verdict on the chemistry side is mirrored on the capital side by a structural problem that capex committees know well: Western project finance time horizons do not stretch to where redox flow scale-up curves bend, and the only project finance that does — Chinese state-anchored long-duration storage finance — sits inside a procurement architecture that most Western capex committees cannot reach.

• Lead-acid stationary BESS. This is a mature, gradient-improvement regime. The global stationary lead-acid storage market sat at $9.4 billion in 2025, with EnerSys, Exide, East Penn, C&D Technologies, GS Yuasa, Narada and Crown Battery the entrenched players. 

The technology roadmap is incremental: thin-plate pure lead, carbon-enhanced AGM, lead-carbon hybrid, bipolar lead-acid. No single step-change qualification gate organises the procurement cycle. 

• Redox flow battery scale-up. This is configuration-driven rather than chemistry-driven. Rongke Power has now commissioned the world’s first GWh-scale vanadium flow project at Jimusaer (200MW/1,000MWh) alongside its 175MW/700MWh Xinhua Power Generation Wushi system, and Invinity has won UK approval for a 20.7MWh deployment. 

IDTechEx forecasts the redox flow battery market reaching $9.2 billion by 2036 at a 27% CAGR. The bottlenecks are pump architecture, membrane cost, electrolyte hedging and tank geometry — variables that scale continuously, not via a discrete chemistry cliff. 

Sodium-ion entry-tier commercialization. This is still in the “insufficient mass-production sample” zone V8 would today flag as Tier D for identifiability reasons. CATL’s Naxtra is moving to GWh-scale industrialization in 2026 at up to 175Wh/kg, with the Changan Nevo A06 — the world’s first mass-production passenger vehicle equipped with sodium-ion batteries — set to reach the market by mid-2026, and the first 300+ Ah sodium-ion BESS cells coming off Envision’s line in March 2026. 

The chemistry will likely fit a cliff in three or four years but today there are not enough disclosed data points to identify one. 

For a Batteries International reader, the implication is direct: these three regimes need a planning architecture that looks nothing like the NMC qualification-gate playbook, and pretending otherwise is the failure mode V8’s Tier D output is built to surface. 

The discrimination matters for capital allocation

A model that finds cliffs everywhere is suspect. A model that finds cliffs in NMC and LFP, suggests one in SSB, and explicitly returns “no identifiable cliff” in memory semiconductors and the Bondalti chlor-alkali membrane series is doing real discrimination. That discrimination is what makes the verdict investible.

The 2027 capex committee in the opening scene was facing exactly this choice. A $500 million high-nickel NMC expansion lives in cliff territory: discrete qualification windows, OEM nomination cycles tied to vehicle platforms, EU Battery Regulation deadlines that hard-gate market access (carbon-footprint declarations for industrial batteries above 2 kWh from February 18, 2026, digital battery passports from February 18, 2027), and procurement contracts that bake in three-to-five-year exclusivity.

The right planning architecture is step-gate, multi-year, and locked. 

A $500 million lead-acid telecom refresh lives in gradient territory: annual recompetition, incremental AGM and lead-carbon upgrades, replacement cycles every 8–10 years across a deployed base, and a supplier set that competes on service network and recycling logistics rather than discrete chemistry cliffs. 

The right planning architecture is rolling, scenario-based, and recompete-friendly. 

Same dollar amount. Different decision rule. V8’s Tier-D-as-output structure is what lets a committee tell them apart on the record rather than after the fact. 

THREE ACTIONS FOR SENIOR BATTERY INDUSTRY DECISION-MAKERS 

One — tag every active capex line as cliff-regime or gradient-regime, and budget the planning cycle accordingly. 

Where V8 (or a comparable public-disclosure cliff test) returns a Tier A verdict, run a step-gate Gantt with multi-year qualification windows. 

Where it returns Tier D or sits in structurally Tier-D-shaped territory, compress the planning cycle to annual or semi-annual reviews. 

Quantitative effect: cliff-regime programs gain six to 12 months of advance positioning around qualification windows; gradient-regime programs shave one to two quarters off internal scenario-planning cycles by dropping unwarranted multi-year locks. 

Qualitative effect: ends the failure mode where a single Gantt chart is run across structurally different capex lines.

Two — reallocate roughly 10%–20% of the screening-stage capex pool from “single playbook” to “tier-aware” assessment. 

A program manager weighing an NMC expansion against a lead-acid retrofit, a vanadium flow scale-up or a sodium-ion BESS pilot should be required to attach a regime tag — cliff or gradient — to each proposal, with a one-page V8-style rationale. 

Quantitative effect: the IRR delta between correctly-classified and mis-classified projects in 2024–2026 vintages is typically 200–400 basis points; the due-diligence cost of adding a tier tag per target is trivial by comparison

. Qualitative effect: lifts the screening-stage signal-to-noise ratio without lengthening the cycle. 

Three — require negative-finding disclosure in supplier and partner technology roadmaps.

This is where a vendor claims a cliff that V8-style analysis cannot identify, treat that as a yellow flag — not because the vendor is wrong, but because the underlying disclosed data does not yet support the claim. 

Batteries International readers commissioning RFPs for stationary BESS, flow systems, sodium-ion pilots and bipolar lead-acid can specify “Tier D-equivalent honest reporting” as a procurement requirement at near-zero cost. 

Quantitative effect: no upfront cost; reduces post-award rework where vendor claims fail to materialize. 

Qualitative effect: builds counter-pressure against the publication-bias asymmetry that currently favours over-confident vendor narratives. 

The investment committees that will look disciplined in 2030 are the ones who, in 2026, separated their cliff-shaped capex from their gradient-shaped capex on the record — not the ones who ran a single Gantt chart across both and discovered the difference the hard way. 

Sang Bong Song is principal analyst at Song Industry Research, Seoul. He authored the open-access IDP V8 framework (Zenodo, 2026); earlier work appeared in SemiWiki and Battery Power Online.

Image credit: Shutterstock