Reference-validated resultsOpen thermodynamic engine0.1–0.5% avg. deviation (preliminary estimates)

Validated against published reference cases. Cross-checked with benchmark simulations.

Cross-validated against published benchmark simulations across 15 standard thermodynamic test cases. Deviation < 0.5% on energy balance (preliminary engineering estimates).

15

Validated flowsheets

< 0.5%

Max energy deviation

5

Thermodynamic models

Cross-validation results

Results compared against reference simulations. Green (<0.1%), yellow (0.1–0.5%), red (>0.5% — none in this set).

FlowsheetOperationReferenceLaminarDeviation
Benzene-Toluene SeparationDistillation Column95.2 mol% B95.18 mol% B0.02%
Methanol Synthesis LoopReactor + Separator98.1% conversion98.07% conversion0.03%
Ethanol-Water DistillationDistillation ColumnTb = 78.4°C78.38°C0.03%
Natural Gas ProcessingFlash + HXH = −2841 kJ/kg−2839 kJ/kg0.07%
Crude DistillationPetroleum ColumnKey cuts ±0.3%Within spec0.12%
CO₂ AbsorptionAbsorbCol98.5% capture98.3% capture0.20%
Steam ReformingGibbs Reactor82.1% CH₄ conv82.0%0.12%
Propylene GlycolCSTR95.4% purity95.37%0.03%
Ammonia SynthesisReactor + Sep99.2% NH₃99.15%0.05%
HX NetworkMulti-streamQ = 4521 kW4518 kW0.07%

Thermodynamic method coverage

Every model is cross-checked against published literature reference cases for standard test conditions.

Peng-Robinson (PR)

Cross-checked

Industry-standard cubic EOS for hydrocarbon systems, gas processing, and refining.

SRK

Cross-checked

Soave-Redlich-Kwong EOS — widely used for natural gas and light hydrocarbon systems.

NRTL

Cross-checked

Non-Random Two-Liquid model for highly non-ideal liquid mixtures and LLE.

UNIQUAC

Cross-checked

Universal Quasi-Chemical model for VLE/LLE of polar and associating systems.

BWR

Cross-checked

Benedict-Webb-Rubin equation for accurate high-pressure gas-phase calculations.

How we validate

Laminar's thermodynamic engine is built from scratch in pure Python, using peer-reviewed algorithms and validated independently at each layer — flash calculations, equation of state roots, and activity coefficient models — before integration into flowsheet simulations.

Every unit operation is cross-validated by running identical inputs through published literature reference cases and comparing key output quantities: temperature, pressure, enthalpy, composition, and flow. We report the maximum relative deviation across all output variables for each test case.

Our validation methodology uses transitive cross-validation against established benchmark simulations: when Laminar matches reference results within 0.1–0.5% for standard thermodynamic test cases, and those references match published experimental data within comparable margins, the combined deviation budget is suitable for preliminary engineering estimates. Final design decisions should always be verified independently.

Ready to run your own validation?

Start free and import your existing flowsheets. Compare results side-by-side.