Fenske Equation Calculator
Compute the minimum number of theoretical equilibrium stages required for a binary distillation separation using the Fenske equation. The "best case" stage count at total reflux — actual columns need more stages because they operate at finite reflux.
Last updated: May 2026
Compare with similar
About this calculator
The Fenske equation gives the minimum number of theoretical stages N_min needed to separate a binary mixture from a feed composition to specified distillate and bottoms purities, assuming infinite (total) reflux. The formula: N_min = log[ (x_D / (1 − x_D)) · ((1 − x_B) / x_B) ] / log(α), where x_D is the mole fraction of the more-volatile component in the distillate, x_B is the mole fraction in the bottoms, and α is the relative volatility of the more-volatile to less-volatile component (assumed constant across the column — a reasonable approximation for many binary systems). N_min is the absolute lower bound on stage count; real columns operating at finite reflux always need more stages. Variables: x_distillate and x_bottoms are mole fractions of the more-volatile (lighter) component in their respective products; relative_volatility (α) = (y_A/x_A) / (y_B/x_B) at equilibrium, a property of the binary at the operating conditions. Edge cases: α must be > 1 for separation to be possible (α = 1 means the components have identical volatility and can't be separated by distillation); x_distillate must be > x_bottoms (distillate is enriched in the lighter component); both fractions must lie in (0, 1). For α very close to 1 (e.g., separating isomers), N_min becomes very large — typical examples include ethanol/water with α ≈ 1.7 at low ethanol mole fraction (needs many stages), or hexane/toluene with α ≈ 5–8 (needs few stages). The Fenske result combined with the Underwood equation for minimum reflux and the Gilliland correlation for actual stage count vs reflux ratio forms the classic FUG (Fenske-Underwood-Gilliland) shortcut design method for preliminary column design.
How to use
Example 1 — Standard 95/5 separation. You want a distillate of 95% mole fraction (x_D = 0.95) and bottoms of 5% mole fraction (x_B = 0.05) for a binary with α = 2.5. Enter Distillate Mole Fraction = 0.95, Bottoms Mole Fraction = 0.05, Relative Volatility = 2.5. N_min = log[(0.95/0.05) · (0.95/0.05)] / log(2.5) = log[19 · 19] / log(2.5) = log(361) / log(2.5) ≈ 2.557 / 0.398 ≈ 6.42 stages. ✓ Round up to 7 stages minimum at total reflux. A real column at typical 1.3–1.5× minimum reflux would need approximately 1.5–2× this number, so plan on 10–14 actual stages. Example 2 — High-purity separation with low relative volatility. Pharmaceutical-grade separation: x_D = 0.999, x_B = 0.001, α = 1.5 (a difficult separation). Enter 0.999, 0.001, 1.5. N_min = log[(0.999/0.001) · (0.999/0.001)] / log(1.5) = log[999 · 999] / log(1.5) ≈ log(998001) / 0.1761 ≈ 5.999 / 0.1761 ≈ 34 stages. ✓ Round up to 34–35 minimum theoretical stages — and a real column at finite reflux might need 50–80 stages, which is why ultra-pure separations of close-boiling components often switch to alternative methods (extractive distillation, crystallisation, membrane separation) when α is too close to 1.
Frequently asked questions
Why "minimum" stages, and how do real columns compare?
The Fenske equation gives N_min at infinite (total) reflux — the theoretical best case where every drop of overhead vapour is returned as reflux and no product is withdrawn. At total reflux, the operating lines collapse onto the diagonal (y = x), each stage operates at the equilibrium curve's maximum effectiveness, and you need the fewest possible stages. Real columns operate at finite reflux (because you need to withdraw product), which means each stage achieves less separation than at total reflux, and more stages are needed. The Underwood equation gives the minimum reflux ratio R_min for finite stages; the Gilliland correlation relates actual stages to actual reflux for any R > R_min. As a rule of thumb, designers run at R = 1.2–1.5 × R_min and accept about 2× N_min actual stages. So Fenske gives the absolute lower bound; doubling it gives a rough engineering estimate of real stage count for typical designs.
What is relative volatility and where do I get it?
Relative volatility α is defined as α_AB = (y_A/x_A) / (y_B/x_B) at vapour-liquid equilibrium, where x and y are mole fractions in liquid and vapour respectively. For ideal solutions following Raoult's law, α equals the ratio of pure-component vapour pressures: α_AB = P°_A / P°_B. For real (non-ideal) mixtures, α depends on composition and temperature; for distillation calculations the value at average column conditions is used. Sources: experimental VLE data (DECHEMA, NIST WebBook), thermodynamic models (NRTL, UNIQUAC, Wilson, UNIFAC for predictions), and Antoine equation for pure-component vapour pressures. Typical values: methanol/water α ≈ 7.5–4 (depending on composition); ethanol/water α ≈ 1.7–1.0 (azeotrope at ~89 mol%); benzene/toluene α ≈ 2.4; n-hexane/n-heptane α ≈ 2.5; ortho-xylene/para-xylene α ≈ 1.1 (very difficult); air separation O₂/N₂ α ≈ 4.
What happens at an azeotrope?
An azeotrope is a composition at which the vapour and liquid phases have the same mole fraction — α = 1 locally — so simple distillation cannot enrich beyond that point. Ethanol/water is the classic example: at 89 mol% (95.6 wt%) ethanol the azeotrope blocks further enrichment by ordinary distillation, which is why anhydrous ethanol requires alternative methods (azeotropic distillation with benzene historically, molecular sieves now, or pressure-swing distillation). The Fenske equation produces nonsense if you try to push x_D past the azeotrope composition — you'd need infinite stages. For azeotropic systems, design either targets a product on the favourable side of the azeotrope, uses an entrainer to shift the azeotrope position, or switches technology entirely. Pressure-swing distillation exploits the fact that some azeotrope compositions shift with pressure (THF/water, MEK/water, etc.).
What are the most common mistakes using the Fenske equation?
The first is forgetting that N_min counts theoretical equilibrium stages, not actual trays — real trays achieve only 40–80% of equilibrium (the Murphree efficiency), so actual tray count = N_theoretical / efficiency. The second is using a relative volatility from the wrong temperature or pressure; α typically decreases with increasing temperature, and column conditions usually need a value averaged across the column. The third is applying the equation to azeotropic systems where α varies dramatically with composition; the constant-α assumption breaks down and predictions are wildly wrong. The fourth is treating Fenske's result as a real design rather than a starting point — actual columns need finite-reflux design (FUG shortcut) or rigorous stage-by-stage simulation (RadFrac in Aspen, etc.) to size properly. The fifth is interpreting low N_min as "this separation is easy" without checking the required reflux ratio — low α means low N_min only if you're willing to run at very high reflux, with huge energy cost.
When should I not use this calculator?
Skip it for multi-component distillation (more than two species) — multi-component shortcut methods (Fenske-Underwood-Gilliland) need extension to handle key components and non-key distribution. Don't use it for azeotropic or highly non-ideal mixtures where α varies strongly with composition; use rigorous stage-by-stage simulation with the appropriate thermodynamic model. It's the wrong tool for design of actual columns beyond preliminary stage count — full design requires finite-reflux analysis (Underwood for R_min, Gilliland for stage vs reflux), tray hydraulics, pressure drop, condenser/reboiler sizing, and rigorous simulation. Avoid it for batch distillation where compositions change with time; the steady-state Fenske doesn't apply. Finally, don't use it for reactive distillation, distillation with phase changes (extractive), or any non-equilibrium separation; specialised methods are needed.