English is currently available.
Chemical Engineering DISCUSSION
I am doing a McCabe-Thiele design for a binary mixture with a relative volatility of 2.5. The feed is 50 mol% of the light component as saturated liquid, and I want 95 mol% in the distillate and 5 mol% in the bottoms.
The textbook says to find the minimum reflux ratio and then operate at 1.2 to 1.5 times that value. What does the minimum reflux physically represent, how is it calculated for this case, and why that particular multiple?
Reflux is the condensed overhead liquid sent back down the column. It provides the liquid that the rising vapour is contacted with on each stage above the feed, and the reflux ratio R = L / D sets the slope of the rectifying operating line, R / (R + 1). More reflux moves the operating line away from the equilibrium curve, each stage achieves a bigger composition change, and fewer stages are needed.
The price is energy. Everything that leaves the top as vapour must be boiled in the reboiler and condensed again: V = (R + 1) × D. So reflux ratio trades column height against heating and cooling duty.
The minimum reflux ratio is the value at which the operating line just touches the equilibrium curve. At that pinch point the driving force is zero and an infinite number of stages would be needed; below it the separation is impossible however tall the column.
For a saturated liquid feed the pinch is at the feed composition. With x = 0.5, the equilibrium vapour is y = α × x / (1 + (α - 1) × x) = 2.5 × 0.5 / 1.75 = 0.714. Then R_min = (x_D - y) / (y - x) = (0.95 - 0.714) / (0.714 - 0.5) = 1.10. This holds for a well-behaved equilibrium curve such as constant relative volatility; in strongly non-ideal systems the line can touch the curve elsewhere first.
The opposite limit is total reflux, where no product is withdrawn and the operating lines lie on the diagonal. That gives the minimum number of stages, from the Fenske equation: N_min = ln[(x_D / (1 - x_D)) × ((1 - x_B) / x_B)] / ln α = ln(19 × 19) / ln 2.5 = 5.89 / 0.916 = 6.4 theoretical stages, counting the reboiler as one.
Every real design lies between these two limits: more than 6.4 stages and more than 1.10 reflux ratio. Stepping off stages on the diagram at your chosen reflux tells you where.
The multiple comes from cost. Just above R_min the number of stages falls very steeply as reflux increases, so a little extra energy saves a great deal of column height. Further out the curve flattens: more reflux buys few stages, while reboiler and condenser duty, and the column diameter needed to carry the vapour, keep rising in proportion to R + 1. Total annual cost has a shallow minimum, which for most systems lies between about 1.1 and 1.5 times R_min.
For your case at 1.3 times, R = 1.43. A mass balance on 100 kmol/h of feed gives D = 100 × (0.5 - 0.05) / (0.95 - 0.05) = 50 kmol/h, so the vapour load is V = 2.43 × 50 = 122 kmol/h. When energy is expensive the optimum moves towards the lower end of the range.
Two practical points for the design. The feed condition changes the answer: a partly vaporised feed tilts the feed line, moves the pinch and raises R_min, so recalculate if the feed is not saturated liquid. And theoretical stages are not trays. Divide by the tray efficiency, which depends on the system and tray type and is often somewhere between 50 and 80 percent, to get the number of real trays.
Operating a little above the economic minimum is also deliberate. A column designed right at the edge has no room to hold product purity when the feed composition drifts, whereas some spare reflux gives the control system something to work with.
Back to Chemical Engineering community
