01 / THE CHOICE BEFORE THE FLOWSHEET
Your first click already decided the answer.
A property method is a claim about your mixture. Use the tree. Inspect the curve. Then decide whether the claim holds.
A starting point for model selection. Validate against your mixture, conditions and property data before design.
BUILDING REFLUX FOR YOUR WORK
Keep the assumptions visible.
Join the Reflux trial waitlist and tell us which simulator you use. The guide below is free to use without signing up.
Already using Aspen Plus on Windows? Try Reflux Student →02 / FOLLOW THE PHYSICS
What kind of mixture is this?
Choose the description closest to your case. Each branch is a candidate family, followed by the evidence it needs.
POLAR, NON-ELECTROLYTE↓↓
Activity-coefficient family
NRTL · UNIQUAC · Wilson*
YOUR NEXT CHECK
Do the pairs have evidence?
Inspect every relevant binary pair, parameter source and fitted range. For liquid–liquid splitting, use a model that can represent it; Wilson is unsuitable for LLE.
Compare predicted VLE / LLE and any design-critical properties with independent measurements at relevant conditions.
The often-quoted “below about 10 bar” is a screening guideline for activity-based methods, not a universal boundary. Near-critical conditions and model-specific vapour treatments matter. Adapted from Carlson’s selection framework.
03 / SAME MIXTURE. DIFFERENT ASSUMPTION.
The crossing changes the job.
At an azeotrope, liquid and vapour have the same composition. Ordinary distillation at that pressure cannot cross this composition barrier.
Ideal assumptionNonideal behavioury = x
Liquid 35%Vapour 44%Vapour is richer
Qualitative curves, not a simulator result or fitted dataset. The crossing location is based on the reported acetonitrile–water azeotrope near 0.7 mole fraction at 1 atm. See Carlson, Fig. 4 for the temperature–composition data.
04 / THE EMPTY CELL
Blank does not mean verified.
PAIR i ↔ j?
A missing value can be estimated, defaulted, warned about or rejected. The behaviour belongs to the simulator and the selected model.
If a selected model falls back to unit activity coefficients, it has made an ideal-liquid assumption. COCO/TEA’s documentation defines ideality as γᵢ = 1; that definition is not evidence that every simulator handles a missing pair this way. COCO activity-model documentation ↗
A zero cubic-EOS interaction parameter does not make the entire mixture an ideal gas. Check what the parameter means before interpreting the number.
05 / BEFORE YOU TRUST THE STAGES
Make the model earn the answer.
- Define the job. Which phases and properties control the decision?
- Record the envelope. Components, composition, temperature and pressure.
- Inspect the pairs. Source, fitted range, missing data and warnings.
- Plot independent evidence. Compare the relevant VLE, LLE or property data.
- Recheck the equipment. Stage count also depends on reflux, specifications and efficiency. Height does not come from a curve alone.