TEA Handbook

Concept

cross

TEA Readiness Levels (TEA-RL)

The TEA Readiness Levels (TEA-RL) are a four-level ladder describing how far a techno-economic model has been built out — from a one-line process sketch to a diligence-grade model — with each level fixed by the depth of its scope, the quality of its data, and the analysis it can support. It grades the analysis itself, distinct from the technology behind it (not a Technology Readiness Level) and from the error bar on the answer (its accuracy class). It is a handbook-specific framework, not an industry standard like the FEL/AACE estimate classes.

What it is. A TEA is built up layer by layer, in increasing depth, precision, and accuracy; TEA-RL names four resolved stops along that path. Each level is set by three things: the objective it serves, the key question it can answer, and the concrete contents of the model at that stage — its process flow, cost build, cost data, sensitivities, visuals, and scenario handling. The same model is a different level at different stages of build-out, even for an identical technology.

LevelObjectiveKey question it answersOwner
1Outline the basic process conceptWhat unit operations does this require?Founders
2Establish a first-pass, end-to-end economic pictureIs there an economically viable concept here?Founders
3Build a robust model for internal decisions and optimizationWhich unit operations and assumptions need more granularity?Founders + engineer
4Create a high-fidelity model suitable for external diligenceWhich technical or strategic levers most improve feasibility?Founders + engineer

What deepens between levels. The ladder isn’t about polish — it’s about resolution. Each row below sharpens as the level climbs:

DimensionLevel 1Level 2Level 3Level 4
Model footprintBlock-flow sketch~2 connected sheets~4–7 sheets, 2–3 deep-dives~8–12 sheets, heavy validation
Process flowBlock-flow diagramAll unit ops in one sheet, ~4–5 lines each~8–10 lines each; core-tech deep-divesAdds auxiliaries (utilities, wastewater)
Data inputDesktop refs for a few key assumptionsRough estimates / comps for all assumptionsDetailed ranges, expert-validated, some quotesMultiple vendor quotes; dedicated validation tabs
CAPEXOrder-of-magnitude equipment list~2–3 priced items per unit op → total capex; annualized via CRFSizing tied to flow; 2–3 refs per major driver; plant- and unit-level comps5+ refs for the ~10 biggest drivers; vendor quotes for core tech
Variable costs & pricingDatapoints for 2–3 main inputs/outputsAll costs linked to the mass & energy balance; incl. consumablesReferenced ranges for all main inputs/outputsValidated, location-specific pricing; offtake pricing
Fixed costsRough % of capex or revenueBottom-up labor, maintenance, overheadDetailed staffing plan; taxes, insurance
SensitivitiesRough (waterfall + tornado)Univariate sweep across a comprehensive variable setUnivariate + multivariate; scenarios
VisualsMarket-sizing waterfallUnit-economics waterfall; tornado chartComps, cost curvesAdvanced analysis visualizations
ScenariosSingle base caseBase caseDashboard over a defined set (e.g. high/low price); feedstock/output togglesIntegrated scenario manager; version control + changelog

Relationship to accuracy class. TEA-RL and accuracy class read the same maturity axis from opposite ends. TEA-RL describes the model you built — how resolved its scope, data, and analysis are; accuracy class describes the ± band on the answer that resolution earns. They move together: a Level-1 outline supports only an order-of-magnitude (FEL-1 / AACE Class 5) result, while a Level-4 diligence model built on vendor quotes can carry a far tighter class. The level is the input maturity; the class is the output tolerance.

Distinct from a Technology Readiness Level. A TEA-RL grades the economic analysis; a Technology Readiness Level grades the technology. The two are independent: a bench-scale process (low technology readiness) can carry a thoroughly built Level-3 model, and a commercially mature technology can sit behind a Level-1 back-of-envelope. The shared “readiness level” name is the only thing the two scales have in common.

Limits & typical error

See also

  • Uncertainty & accuracy class — the output-side ± band that a given build level can honestly earn.
  • Tornado chart — the one-way sensitivity analysis that first appears at the middle of the ladder.
  • False precision — reporting a result to finer resolution than the model’s level supports.