Research study · Case study
Water–ethanol VLE at 100 kPa
Cross-engine validation of BioSTEAM/ThermoSTEAM and DWSIM against a frozen 26-point experimental water–ethanol vapor–liquid-equilibrium dataset at 100 kPa.
Claims on this page apply only to the stated system boundary, methods, source revision, and evidence status.
- Source revision
- nonideal-vle-water-ethanol-100kpa-v1.1 · formal run 20260915T182649Z
- Last verified
- 2026-09-15
Decision question
Can two independent open process-simulation paths reproduce the published 100 kPa water–ethanol VLE trend and azeotropic region within preregistered error thresholds?
System boundary
Binary water–ethanol vapor–liquid equilibrium at 100 kPa. The benchmark covers equilibrium temperature, vapor-phase ethanol composition, and the azeotropic region for the frozen published dataset; it does not generalize to other mixtures, pressures, property methods, or unit-operation models.
Functional or reference basis
Twenty-six experimental equilibrium points reported at 100 kPa, plus a frozen azeotrope search grid.
Methodology
The study froze the experimental data and acceptance thresholds before execution. BioSTEAM/ThermoSTEAM used its qualified thermodynamic path. DWSIM 10.2.5 used Modified UNIFAC (Dortmund) and its native pressure-and-vapor-fraction material-stream flash at P=100000 Pa and V=0. The corrected revision changed only the DWSIM query path after the first formal method was shown to be confounded.
Assumptions
The published experimental dataset is treated as the reference. Each engine is evaluated through the stated qualified runtime and query path. The comparison is between model predictions and the frozen dataset, not between proprietary plant measurements or customer models.
Results
BioSTEAM/ThermoSTEAM achieved temperature MAE 0.22664 K and vapor-ethanol-composition MAE 0.008735, with an azeotrope estimate near x=0.89404 and T=351.06668 K. DWSIM 10.2.5 achieved temperature MAE 0.20198 K and vapor-ethanol-composition MAE 0.006998, with the formal grid location near x=0.890 and T=351.26868 K. Both paths passed the frozen acceptance criteria. All 55 DWSIM states in the formal run solved without calculation errors.
Interpretation
Within this narrow binary system and the stated methods, both engines reproduce the published equilibrium data closely enough to pass the preregistered benchmark. The result supports the qualified thermodynamic query paths; it does not establish broad parity between the engines.
Limitations
DWSIM returned y=x at several adjacent grid points near the azeotrope, so the reported x=0.890 location is a grid-resolution location within a numerically flat region rather than a uniquely resolved mathematical root. The original v1.0 failure remains preserved as historical evidence and is not erased by the corrected v1.1 qualification.
Reproduction
Use the frozen 100 kPa data from Susial Badajoz et al. (2021), the qualified BioSTEAM/ThermoSTEAM runtime, and DWSIM 10.2.5 with the preregistered pressure-and-vapor-fraction bubble-point query. Recompute temperature and vapor-composition errors point by point and rerun the azeotrope grid.
Review the underlying methods.
Use the documentation to inspect the tools and workflows behind this study.