1-805-484-2992
Quotes - Engineering - SalesDephlegmators
Column-Mounted Partial Condensers for Internal Reflux Generation
A dephlegmator sits on top of the column and makes its reflux internally — no drum, no pump, no return piping. It buys simplicity and containment at the cost of control authority. We will help you decide whether that trade is right for your column.
A dephlegmator is a partial condenser mounted directly on top of a distillation column. Vapor from the column passes up into the tubes, coolant flows on the shell side, part of the vapor condenses, and the condensate falls straight back down into the column as reflux. Uncondensed vapor continues out the top as product.
It is a familiar piece of equipment in more places than people expect. Vertical shell-and-tube arrangements with coolant on the shell side and feed vapor in the tube bores are the common industrial form; vertical brazed aluminum plate-fin dephlegmators are used in cryogenic distillation for gas separation in refineries and petrochemical plants. In craft distilling the same device sits at the top of a column still, where varying the coolant flow changes the reflux ratio and therefore the proof and character of the spirit.
A partial condenser mounted directly on top of a distillation column. Vapor from the column passes up into the tubes, coolant flows on the shell side, part of the vapor condenses, and the condensate falls straight back down into the column as reflux.
What a dephlegmator gives you. Considerably less infrastructure — no reflux drum, no pump, no return piping, less structural steel and no plot space for any of it. A much smaller inventory of material at the top of the column. Fewer flanged joints and rotating seals, so fewer potential leak points. No need to convey liquid reflux back up to the column at all. On processes handling hazardous, toxic or corrosive substances, those last points are a genuine safety advantage rather than merely a cost saving.
What it costs you. Control authority, primarily. An external condenser with a reflux drum lets you measure and manipulate reflux flow directly, and the drum's inventory buffers the column against disturbances in inert loading, vapor rate, coolant flow or coolant temperature before they reach the trays. A dephlegmator has essentially no liquid holdup, so reflux can only be inferred from an energy balance, and any disturbance in condensing rate propagates into the column immediately. On columns with tight product specifications or frequent rate changes, that sensitivity can be a serious operating problem — and it is the reason dephlegmators are not simply the default choice despite their cost advantage.
Mechanical realities. The unit has to be mounted and structurally supported at the top of the column. Coolant must be piped up to that elevation and supplied there. Maintenance means physically removing the unit, and cleaning is more awkward than on an exchanger at grade. Where fouling is expected, this matters a great deal.
Countercurrent flow and flooding. vapor rises while condensate drains down the same tubes, so the tubes must be designed for countercurrent flow, and flooding imposes a firm limit on vapor velocity. Most vertical units in this service are designed for downflow condensation with a deliberate margin against that limit. A useful practical guide is that the dephlegmator's tube bundle should have roughly enough open area to approximate the column's own cross-section — a badly undersized bundle chokes the column regardless of how much surface area it contains.
Staged designs. Some arrangements combine sections: a shell-and-tube or plate-fin upper section where external coolant does the work, over a packed lower section where heat exchanges directly between rising vapor and falling condensate. Because only the vapor condensing at the very top must be chilled to the lowest temperature, and the condensate itself performs part of the cooling duty on its way down, this can use noticeably less refrigeration than simple partial condensation.
Related pages: reflux condensers, distillation condensers, overhead condensers, and distilled spirits condensers.
Column-top questions
It is a partial condenser whose condensate drains back into the column by gravity, best pictured as a few extra separation stages sitting on top of the column rather than a utility exchanger. Because the heavier components condense first on the way up through the tubes and the lighter ones carry on as vapor, a dephlegmator is doing fractionation, not just heat removal, and a well designed unit is credited with between one and three theoretical stages in the column design.
The name comes from an old distiller's term for removing the 'phlegm', the watery fraction, from a spirit. In practice the output is controlled by the temperature of the vapor leaving the top: raise coolant flow and that temperature falls, more condenses and returns, and the product that gets through is lighter and purer.
The savings are concrete items: a reflux drum, a reflux pump and its installed spare, their level and flow loops, two piping runs to grade and back, and the steel that holds it all. On a 3 ft diameter column that package can cost as much as the condenser itself, and it needs plot space beside the column that a crowded unit may not have.
The second advantage is harder to price but often decisive: the liquid inventory at the top of the column drops from the hundreds of gallons held in a drum to the thin film on the tubes, so a hazardous overhead has far less to release if a seal or a gasket fails. In craft distilling there is a third benefit, which is that coolant flow becomes a direct and responsive handle on reflux ratio and therefore on the spirit's proof.
You give up direct control of reflux, and on a dephlegmator every upset in the coolant reaches the trays immediatly because there is no drum to absorb it. A 3 F swing in cooling water temperature, which a tower loop produces every afternoon, changes the condensing rate and therefore the reflux ratio within seconds, and the column has to ride it out.
The usual mitigation is a tempered coolant loop: a circulating pump and a three-way valve blend return water with supply so the dephlegmator sees a steady temperature regardless of the tower, and the controller trims that temperature against the vapor outlet temperature. That restores a good deal of control but adds back a pump and a loop, which eats into the simplicity argument. Where product specifications are tight and feed rates swing, an external condenser and drum is still the safer choice.
A dephlegmator is designed to let a defined fraction of the vapor through as product, so its sizing is a fractionation calculation, while a reflux condenser is designed to return everything and its sizing is a flooding calculation. That difference shows up in the hardware. A dephlegmator tube bundle has to be rated against the changing composition along its length, with the condensing temperature falling as the vapor gets lighter, and the vapor outlet temperature is a controlled variable.
A knock-back unit on a reactor sees nearly constant composition and is rated for the peak vapor rate with a margin below the flooding velocity; its outlet is only inerts. The two are often confused because both are vertical, both drain by gravity and both sit on the vessel they serve. The question to ask is whether vapor is supposed to leave the top at all.
Size it from the flooding velocity first, then check the result against the column cross-section; the two should land close together. Work it in that order: from the peak vapor rate and the vapor density, pick a tube velocity with a comfortable margin under the countercurrent flooding limit, and the tube count falls out. Larger tubes help more than people expect, because the flooding velocity rises roughly with the square root of the tube diameter, so moving from 1 in to 1.5 in tubes allows about 20 percent more velocity per tube as well as more area per tube.
If the bundle that results has an open area well short of the column's, the design is relying on surface that the vapor cannot reach without flooding, and the answer is a larger shell, not more tubes in the same one. Headroom above the column then becomes the practical limit on tube length.
Yes, and on most columns the vapor outlet temperture is the only measurement worth controlling on, because it tracks the composition of what is getting through far more directly than coolant flow does. The loop is simple: a temperature element in the vapor line just above the dephlegmator sets the coolant valve, and the setpoint is the dewpoint of the product you want, read from the vapor-liquid equilibrium data at the column pressure. For an ethanol still at atmospheric pressure, holding the vapor leaving at 174 F keeps the product near azeotrope strength, around 190 proof, while a setpoint of 180 F lets more water through.
Two cautions apply. Put the element close to the outlet, since a lag of even a minute makes the loop hunt. And give the coolant valve a usable range, because a dephlegmator at the low end of its coolant flow behaves nonlinearly and a valve cracked 5 percent open controls badly.
