Wiped-Film and Short-Path Rolled-Film Evaporation

What changes mechanically industry practice, not published data

A short-path still heats a static pool of liquid in a flask and waits for the volatile components to leave it. A wiped-film evaporator does the opposite: it feeds liquid continuously onto the inside of a heated cylindrical wall, where a rotating assembly of wiper blades or rollers spreads it into a thin, constantly renewed film and pushes it down the wall as it travels. The film is thin enough — a fraction of a millimetre — that a molecule anywhere in it is close to both the heated surface and the vapour space, so the heat transfer coefficient is high and the diffusion path to the surface is short. Vapour leaves the film, travels a very short distance to an internal condenser, condenses, and runs down to a distillate outlet. Whatever did not evaporate keeps travelling down the wall and leaves as residue at the bottom. Nothing accumulates, nothing sits. The mechanical agitation also means the unit tolerates a viscous, fouling feed that would bake onto a static flask wall, because the wipers are continuously scraping the surface clean.

• Thin film plus mechanical renewal equals very high heat-transfer coefficient at low temperature difference.
• Continuous in, continuous out: two product streams and no batch cycle.
• The wipers keep the heated surface clean, which is what allows viscous and fouling feeds.

Sources: Green DW 2019* · Composite: extraction-equipment vendor documentation 2026*

Residence time and heat history industry practice, not published data

This is the whole point. In a batch flask, the last molecule to distil has been sitting at temperature for as long as the run has been going — for a multi-hour pass, that is hours of thermal exposure for material that spends one second actually vaporising. In a wiped-film unit, every molecule experiences essentially the same short exposure, measured in seconds, because the film is moving and being replaced. The integral of temperature over time — the quantity that actually drives isomerisation and oxidation, as set out on short-path-distillation — is therefore smaller by orders of magnitude per unit of product, even though the wall temperature may be the same or slightly higher. The observable consequences are lighter colour at equal potency, lower CBN generation for the same throughput, and a much narrower spread of quality across a production day, since there is no start-of-batch versus end-of-batch difference. The trade is that the separation achieved in a single pass is less sharp than a carefully fractionated batch run, because a wiped-film stage is closer to a single-stage flash than to a fractionating column: you get a clean split between two streams, but you do not get the fine sequence of narrow cuts that a patient operator can take off a cow.

• Uniform, short thermal exposure for every molecule instead of a distribution running from seconds to hours.
• Lighter colour and lower degradation markers at equal potency.
• Less sharp fractionation per pass — sequential stages or multiple units are used to get the cuts back.

Sources: Green DW 2019* · Composite: extraction-equipment vendor documentation 2026* · Fairbairn JW 1976*

Internal condenser geometry industry practice, not published data

Most cannabinoid wiped-film units place the condenser inside the evaporator body, coaxially, a short distance from the film. The reason is conductance. At the pressures where cannabinoids distil, the vapour is in a regime where the mean free path is long and the flow is not a bulk fluid flow but something closer to free molecular travel, so any length of pipe, any bend and any constriction between the evaporating surface and the condensing surface costs pressure and therefore costs you the temperature advantage you built the vacuum system to get. Putting the condenser two or three centimetres from the film all but eliminates that loss and lets the effective pressure at the evaporating surface approach what the pump can deliver. The same logic is why the short-path flask geometry exists; the wiped-film unit simply takes it further. The engineering consequences are that the internal condenser must be designed so condensate cannot drip back onto the film — otherwise you are running an unintended reflux and destroying your separation — and that the coolant temperature on the internal condenser has to be set high enough that the cannabinoid fraction stays fluid enough to run off, but low enough that it does not pass through to the vacuum system.

Sources: Green DW 2019* · Composite: extraction-equipment vendor documentation 2026*

Throughput and why volume forces the move contested industry practice, not published data

Throughput on a wiped-film unit scales with the heated surface area and is quoted per hour rather than per batch, which is the economic difference. Small units used in cannabinoid work are commonly specified at something in the region of 0.5 to 2 kg/h of feed, mid-size units in the several kilograms per hour range, and large industrial thin-film and falling-film trains far beyond that. A short-path batch setup at the scale most small processors own handles a charge of a few litres and takes most of a shift per pass once you include pull-down, fraction changes, cool-down and cleaning, so its effective rate is low and its labour content per kilogram is high. Somewhere in the low single-digit kilograms per hour of sustained demand, the arithmetic flips: the wiped-film unit costs more to buy and to commission, but its output per operator-hour and its consistency across a day make the cost per kilogram lower and the product more saleable to a customer who specifies colour. The exact crossover is a function of local labour cost, how much your market pays for colour and terpene integrity, and how much idle time your batch setup actually has, so it is a calculation each operation does for itself rather than a published constant.

Contested — caveat. Throughput figures for cannabinoid wiped-film units are vendor specifications quoted against unstated feed compositions, and real sustained rates on winterized cannabinoid crude are frequently well below nameplate. The switch-over volume is an operation-specific economic calculation, not an industry constant. Treat every number in this section as an order-of-magnitude orientation.

Sources: Composite: extraction-equipment vendor documentation 2026*

The trade-offs, stated honestly industry practice, not published data

A wiped-film unit is not a better short-path still, it is a different machine with a different failure surface. Capital cost is several times that of a comparable batch setup once the chiller, the vacuum train, the feed pump, the heated feed vessel and the discharge handling are included. There are more variables to tune and they interact: feed rate, wall temperature, wiper speed, internal condenser temperature, vacuum, and the temperature of the feed as it arrives all move the split, and a change in one requires re-trimming the others, so commissioning is a real project rather than an afternoon. The feed must be prepared: decarboxylated, winterized, filtered, and — critically — degassed and solvent-free, because a continuous unit has no patient pull-down phase in which volatiles can be stripped, and solvent-wet feed arriving at a hot wall under deep vacuum flashes, floods the vapour space, carries liquid over into the distillate and can pull the vacuum out from under the whole process. That is why most operations put a dedicated solvent-stripping or devolatilisation stage ahead of the wiped-film stage rather than trying to do both in one unit. Maintenance is also different in kind: wipers wear, seals on a rotating shaft under deep vacuum are a maintenance item and a leak path, and a mechanical seal failure is both a vacuum fault and a contamination route.

• Higher capital cost, and the ancillaries are a large part of it.
• More interacting variables; commissioning is a project, not a setup.
• Feed must arrive decarboxylated, winterized, filtered, degassed and dry — a stripping stage upstream is normal.
• Rotating-shaft vacuum seals and wiper wear are recurring maintenance and a leak path.
• Less forgiving of a mistake: a bad feed ruins a continuous run in seconds rather than over a batch you could have caught.

Sources: Composite: extraction-equipment vendor documentation 2026* · Green DW 2019*

Short path or wiped film: making the decision industry practice, not published data

The decision is mostly about volume, with product positioning and staffing as the secondary terms. The table gives the axes as they actually bite.

Sources: Composite: extraction-equipment vendor documentation 2026* · Green DW 2019*

Safety, as part of the operation

A wiped-film unit adds rotating machinery and a hot, pressurised-thermal-fluid circuit to the vacuum and thermal hazards already described for short-path work, and it removes the glass. Interlock the drive so the wiper assembly cannot be energised with the unit open, and treat the shaft coupling and any inspection port as a machine-guarding question. The heating jacket is usually served by a hot-oil circulator running at up to about 200 °C: that circuit has its own expansion, relief and leak hazards, hot oil causes severe burns, and an oil leak onto a hot surface is a fire. Thermal-fluid hoses and fittings are a consumable, not permanent fixtures. Because the unit is continuous, an upset develops while material is still being fed, so the feed pump needs to be interruptible from where the operator stands, and the control scheme should stop feed on loss of vacuum, loss of coolant or over-temperature rather than relying on someone noticing. Solvent-wet feed is a genuine hazard here and not merely a quality problem: a flash in the vapour space can overpressure a vessel designed for vacuum, so a relief path is required and feed dryness is verified before commissioning a run, not assumed. The vacuum train, the cold trap and the peroxide-forming-solvent prohibition are exactly as set out on short-path-distillation and apply unchanged.

• Interlock the drive; guard the shaft and every port.
• Hot-oil circuit at up to about 200 °C — burns, leaks onto hot surfaces, and hoses as consumables.
• Feed must be stoppable instantly, and should stop automatically on loss of vacuum, coolant or temperature control.
• Verify feed dryness before a run; a flash in a vacuum-rated vessel needs a relief path.
• Every vacuum, cold-trap and peroxide rule from short-path work carries over.

Sources: National Research Council (US) 2011* · National Fire Protection Association 2024* · Composite: extraction-equipment vendor documentation 2026*