Falling Film Evaporator Selection: Sizing and Limits

Last updated: 2026-11-05

A falling film evaporator is the right selection when a clean, non-scaling liquor below about 50 cP can keep every tube wetted at 0.1-0.3 m3/h per meter of tube perimeter (0.13-0.4 gpm/ft), and the wrong one when it cannot. This guide sizes the machine from wetting rate, evaporator LMTD, and vapor body velocity, with a 20 m3/h (88 gpm) sodium sulfate example inside the wider evaporation and crystallization equipment hub.

Key Takeaways

• Select a falling film evaporator for clean liquors below about 50 cP with stable tube wetting.
• Design minimum wetting rate: 0.1-0.3 m3/h per meter of tube perimeter for water-like liquors.
• Keep 8-25 K of useful temperature difference per effect; below 8 K, area cost outruns steam savings.
• For scaling brines and crystallizing duties, forced circulation moves boiling off the tube wall.
• Mechanical vapor recompression (MVR) cuts energy to 25-45 kWh per tonne of water evaporated.

What Is a Falling Film Evaporator and Where Does It Fit?

A falling film evaporator is a vertical shell-and-tube evaporator that feeds liquor at the top of a tall tube bundle, lets it run down the inner tube wall as a thin gravity-driven film, and boils it against condensing steam or compressed vapor on the shell side. The distribution plate is a machined, perforated or weir-slotted plate that spreads the feed evenly across every tube opening, and it is the single component on which the whole machine’s performance rests. The vapor body is a wide disengagement vessel that slows the leaving vapor enough for entrained droplets to drop out or be caught on a mist eliminator before the vapor moves to the next effect, the compressor, or the condenser.

The design differs from older natural-circulation types in one decisive way: the boiling happens inside the tube, on the film surface, not against the tube wall in a pool. A calandria evaporator is a short-vertical-tube, natural-circulation evaporator that submerges the tube bundle in a boiling pool several meters deep; the hydrostatic head raises the boiling point at the tube wall, the liquor circulates slowly, and residence time is measured in many minutes. In a falling film unit the holdup per pass is seconds, the film is a few millimeters thick, and the heat-transfer coefficient on water-like liquors reaches 1,200-2,500 W/m2·K — roughly twice what a calandria achieves, which is exactly why the type exists.

Tubeside boiling is the whole point: liquor on the inside of the tubes as a falling film, heating vapor condensing on the shellside. That orientation gives a small holdup, fast response to feed changes, gentle treatment of heat-sensitive solutes, and the ability to run at low temperature difference — the property that makes multiple-effect and MVR arrangements economical. The limit is equally structural. A film only a few millimeters thick cannot tolerate dirty feeds, crystal loads, or a distributor that feeds one tube twice as much as its neighbor, because the starved tube runs dry within seconds and bakes its solutes onto the wall.

The machine therefore lives or dies by two numbers that appear everywhere in this guide: the wetting rate at the top of the bundle and the useful temperature difference per effect. Size both correctly and the evaporator will be the quietest, cheapest-to-run item in the plant; get either wrong and no vendor datasheet will rescue it. For how the area number itself is built up, see our guide to evaporator heat-transfer area.

Falling Film Evaporator Design Parameters and Typical Ranges

The design exercise runs against eight parameters, and the table below is the checklist we work through on every inquiry before any drawing is issued. All ranges are typical and indicative — every value must be confirmed by test work on your own liquor.

The wetting rate is the flow of liquid per unit of tube perimeter at the top of the bundle, and it is the parameter that converts a mass balance into a tube count. Below the minimum, the film breaks into rivulets, patches of wall run dry, and solutes bake on within hours; above roughly three times the minimum, extra flow buys nothing but taller pumps. The number is liquor-specific — surface tension, viscosity, and foaming all move it — so the ranges in the table are starting points, to be confirmed on a short-tube rig.

The evaporator LMTD governs area the way it does in any heat exchanger, with two evaporator-specific corrections. First, the boiling side temperature is not constant down the tube: it rises with hydrostatic head only slightly (the column of liquid is a film, not a pool), but the boiling point rises with concentration toward the tube outlet. Second, BPE eats useful temperature difference invisibly: a liquor boiling 6 K above pure water at the same pressure delivers 6 K less driving force than the steam table suggests. On a sodium sulfate liquor at 25 wt% (wet basis), BPE is typically 2-4 K; on concentrated chloride brines it can exceed 8 K, and that difference alone can add 20% to the required area.

Vapor body sizing follows from vapor volume, not from liquid throughput, and on low-pressure effects the vapor dominates. Water vapor at 60 °C (140 °F) occupies about 7.7 m3 per kilogram, so a duty evaporating 11 t/h (12 short t/h) of water moves nearly 90,000 m3/h of vapor — the body diameter is set almost entirely by that number and by the 2-4 m/s demister velocity limit. Under-sizing the vapor body is the cheapest mistake to avoid on paper and the most expensive one to fix on site, because entrainment carries valuable solute into the condensate and wrecks the condensate quality balance across effects.

When Is a Falling Film Evaporator the Wrong Choice?

Good selection is scenario work: name the scenario, name the winner. Below are the five we meet most often in mining and salt duties, each with an explicit verdict — because “it depends” is not a design deliverable.

Scenario 1: clean, non-scaling liquor, large steady duty. Demineralized-water concentration, purified nickel or lithium sulfate liquors after polishing, and monosodium glutamate-class organics all wet cleanly and never salt. Verdict: falling film is preferred, and at multiple-effect or MVR scale it is normally the lowest lifetime-cost option on the table.

Scenario 2: scaling or salting brine. Sodium sulfate liquors that reach saturation inside the tubes, silica-bearing geothermal brines, and hard make-up water duties all precipitate solids on a hot wall. For liquors that reach saturation inside the calandria, forced circulation is preferred because flashing happens in the separator, away from the tube wall where crystals nucleate and grow. Verdict: falling film stops at roughly 90-95% of saturation; beyond that, hand the duty to a forced-circulation stage — see our forced circulation vs falling film comparison and the companion guide on evaporator scaling control.

Scenario 3: viscous concentrate. Above roughly 50-100 cP the film thickens, the minimum wetting rate climbs, and the heat-transfer coefficient falls toward 800 W/m2·K or worse. For late-stage concentration of viscous liquors, forced circulation or scraped-film equipment is preferred because a film can no longer keep the wall covered at any economical flow. Verdict: falling film for the early effects only, if at all.

Scenario 4: low energy cost is the driver and electricity is cheap. When the duty runs thousands of hours per year and power is affordable, MVR on a falling film effect is the standard low-energy answer: 25-45 kWh per tonne of water against 0.4-0.6 tonnes of steam per tonne for a double-effect. For plants with abundant waste steam and cheap cooling water, multiple effect is preferred because it converts surplus low-grade steam into evaporation with no compressor capex. Verdict: MVR falling film where power is cheap, multiple-effect where steam is nearly free — the arithmetic is laid out in MVR vs multiple effect and in MVR evaporator selection.

Scenario 5: crystallizing duty to dry cake. Any flowsheet that ends in a salt product needs a crystallizer, not an evaporator, at the back end. For duties producing a saleable sulfate salt, a falling film pre-concentrator feeding a dedicated crystallizer is preferred because crystal growth needs supersaturation generated by flashing in a well-mixed body, not on a heated wall. Verdict: falling film for pre-concentration, then a forced-circulation or Oslo-type unit — see crystallizer design and sodium sulfate crystallization.

One structural note that surprises first-time buyers: falling film and forced circulation are not rivals for the whole flowsheet — most real brine plants use both. For high-salinity end stages in particular, our brine concentrator sizing and selection guide shows where the handover point usually sits, and it is normally a saturation line, not a preference.

What Goes Wrong in Service — and How Do You Fix It?

Most selection mistakes on this machine surface in the first months of operation as one of the seven failure modes below. Each is written the way shift reports read: symptom, then cause, then fix.

1. Dry patches and baked-on fouling at the tube wall. Symptom: local dark rings or hard scale visible at the tube sheet during inspection, and U declining within days of a clean. → Cause: maldistribution — the plate installed out of level, orifices eroded or partially plugged, or feed flow pushed below the minimum wetting rate during turndown. → Fix: re-level and re-orifice the distribution plate, verify per-tube flow by a water test, and lock a wetting-rate floor into the control system so turndown can never starve the bundle.

2. Capacity shortfall even with clean tubes. Symptom: evaporation 15-30% below design at full steam pressure. → Cause: film dryout at high ΔT — the designer spent the temperature budget on few effects with large ΔT, and the film broke at the outlet third of the tubes. → Fix: raise top-of-bundle flow with recycle, reduce ΔT per effect by adding an effect or lowering steam pressure, and re-check the BPE assumption against the real liquor.

3. Salting of tubes on sodium sulfate duty. Symptom: hard crystalline layer on the lower tube wall, cleaned only by boiling out with water. → Cause: the liquor crossed saturation inside the tube, and wall boiling nucleated crystals directly on the surface. → Fix: cap the falling film duty at 90-95% of saturation, add a small crystal-free purge, and route the final saturation-to-product step to a forced-circulation crystallizer where flashing makes the supersaturation.

4. Foaming and condensate contamination. Symptom: conductive or organic-laden condensate, demester carryover visible at the inspection port. → Cause: surfactants, residual flotation reagents, or decomposition products stabilizing foam in the vapor body. → Fix: dose antifoam on a controlled basis, enlarge vapor body cross-section or add demister area, and eliminate the reagent source upstream — mechanical fixes alone only move the problem to the next effect.

5. Distribution orifice plugging. Symptom: progressive wetting-rate loss on a subset of tubes, first effect worst. → Cause: upstream grit, pipe-scale, or crystal carryover from a preceding effect lodging in the 3-6 mm orifices. → Fix: install a duplex strainer or filter ahead of each distribution plate, enlarge orifices where the wetting budget allows, and add a periodic full-flow flush sequence at shutdown.

6. Corrosion at the tube sheet and upper tubes. Symptom: pitting or grooving concentrated at the top tube sheet within the first year. → Cause: chloride concentration at the film edge, splashing zones where wetting is intermittent, or a material selected against the feed analysis rather than the concentrate analysis. → Fix: upgrade the affected zone (duplex or higher-alloy tubesheets), verify chloride at final concentration rather than feed, and ensure the wetting floor eliminates intermittent wet-dry cycling, which concentrates chlorides faster than bulk boiling does.

7. Steam consumption above design. Symptom: specific steam consumption 10-20% over the effect-count prediction. → Cause: non-condensable gas accumulation from air in-leakage blanketing the shellside, or unnoticed condensate level covering part of the bundle. → Fix: survey the train with a U-per-effect test, repair leaks and verify venting at each effect’s non-condensable outlet, and check steam trap and condensate level control — these are commissioning items, not design flaws, and they are usually found in the first month.

A pattern runs through all seven: the evaporator rarely fails alone. The chemistry pushed it, the distributor answered, and the fix usually spans both. Diagnose the flowsheet before the tube bundle.

What Does a Falling Film Evaporator Cost to Buy and Run?

Budget-grade planning numbers, with the basis stated so they can be defended in a feasibility review: installed cost of 4,000-9,000 USD per square meter of heat-transfer surface (indicative, 316L stainless, excluding civil works and buildings), which for a 210 m2 (2,260 sq ft) double-effect package gives a 0.9-1.9 million USD equipment-and-installation range, plus vapor bodies, vacuum system, pumps, and controls, taking a complete train to roughly 1.5-3.5 million USD. Alloy moves the number more than size does: duplex 2205 adds roughly 30-50%, and higher nickel alloys can double it. These figures assume a standard 6 m (20 ft) tube design, shop fabrication, and a greenfield installation with normal access — every deviation shifts the range, so treat them as budget-grade only, to be confirmed by quotation.

Energy depends almost entirely on the effect count or the compressor choice, not on the falling film configuration itself.

Assumptions behind the table: clean aqueous duty, condensate returned, 90% availability, and no credit taken for condensate flash steam. At a steam price of 30-60 USD per tonne, the double-effect saving against single effect is worth roughly 15-20 USD per tonne of water evaporated — on our worked example below, that is around 190,000-250,000 USD per year, which is why effect count is argued about in every feasibility study. For a deeper treatment of the trade, see evaporator energy consumption; for the compressor route as a package, see our MVR evaporation crystallization unit, and for the steam-driven equivalent, our triple effect evaporation crystallization unit.

Maintenance is modest by evaporator standards — no circulation pumps on the liquor side in most designs, no high-speed rotating equipment except in MVR — and an indicative budget is 2-4% of purchase price per year (assumption: clean service; scaling service with monthly boiling-out can double it). The distribution plate and demister are inspection items at every shutdown, not consumables; the tubes are the capital at risk, and the wetting-rate discipline in the troubleshooting section above is what protects them.

How Do You Size a Falling Film Evaporator in 7 Steps?

Here is a complete, redoable sizing for a sodium sulfate duty, with every number traceable. The seven steps below are the same ones we run before any quotation.

1. Define the duty and mass balance. Feed is 20 m3/h (88 gpm) of sodium sulfate liquor at 12 wt% Na2SO4 (wet basis), density 1,100 kg/m3, at 40 °C (104 °F). Target outlet 25 wt% (wet basis). Feed mass 22,000 kg/h carries 2,640 kg/h of sulfate; outlet liquor is 2,640/0.25 = 10,560 kg/h, so evaporation is 11,440 kg/h (12.6 short t/h) of water.
2. Choose the configuration. Steam is purchased at 110 °C (230 °F), cooling water allows a 60 °C (140 °F) last effect, and BPE totals about 4 K across two effects. Available ΔT = 110 − 60 − 4 = 46 K, split 23 K per effect — inside the 8-25 K band, so a double effect is feasible without a thermo-compressor.
3. Assign duty per effect. With feed preheated to boiling, each effect evaporates roughly half: 5,700 kg/h. Latent heat at 65-85 °C averages about 2,320 kJ/kg, so thermal duty per effect is 5,700 × 2,320 / 3,600 = 3.7 MW (12.6 MMBtu/h).
4. Fix U and compute area. For a water-like sulfate liquor below 20 wt%, take U = 1,800 W/m2·K (indicative, to be confirmed by test). Area per effect = 3.7 × 10^6 / (1,800 × 23) = 89 m2; allow 10% margin and specify 100 m2 (1,076 sq ft) per effect, 200 m2 total.
5. Convert area to tubes and check wetting. With 6 m (20 ft) tubes of 45 mm (1.8 in) inside diameter, each tube contributes π × 0.045 × 6 = 0.85 m2, so 118 tubes per effect. Effect 1 top flow 22,000 kg/h ≈ 20 m3/h gives 0.17 m3/h per tube, or about 1.2 m3/h per meter of perimeter — four times the 0.3 m3/h·m ceiling-band minimum, comfortably wet. Effect 2 runs at roughly half that, still above the floor.
6. Size the vapor bodies. Effect 2 evaporates 5,700 kg/h of water at about 62 °C, where vapor specific volume is near 7.9 m3/kg: 45,000 m3/h = 12.5 m3/s. At a 3 m/s (9.8 ft/s) demister velocity, body cross-section is 4.2 m2, or a 2.3 m (7.5 ft) diameter vessel — larger for the first effect, which sees the bigger vapor volume.
7. Close the energy balance and cost it. Double-effect steam demand at 0.5 t steam per t water is 5.7 t/h of 110 °C steam; the MVR alternative at 35 kWh per t water draws 400 kW. Run both through the site energy tariff before choosing — the hardware above is identical, and only the energy route differs.

The output: a two-effect, 200 m2 falling film train with 118-tube bundles, 2.3-2.8 m vapor bodies, and an energy decision between 5.7 t/h of steam and a 400 kW compressor — a hardware list a reviewer can check line by line. For bench-scale confirmation of U and wetting behavior on your own liquor, a lab-scale falling film evaporator campaign closes the last data gaps before the drawings are frozen.

Field Data: Three Anonymized Falling Film Duties

Three numbers from projects we have supported, each with its basis stated so the reader can judge transferability.

Sulfate liquor pre-concentration, 316L, first effect. Overall U measured 1,850 W/m2·K at 30 days and 1,650 W/m2·K at 180 days (basis: heat duty divided by LMTD and specified area, computed from steam condensate flow). The decline tracks a 21-day clean-in-place schedule; the duty was kept below 90% of saturation by design.

Sodium sulfate pilot rig, salting onset. On a 20 wt% liquor (wet basis) at 20 K of ΔT, wall salting began within 48 hours whenever the wetting rate fell below about 0.22 m3/h per meter of tube perimeter, and never occurred above 0.30 (basis: replicated 48-hour runs on a 3 m single-tube rig). This is the wetting floor we now apply to similar sulfate duties, and it sits between the table’s generic water-like and viscous bands — which is exactly why liquor-specific confirmation matters.

MVR falling film on magnesium-removal liquor. Specific energy of 31 kWh per tonne of water evaporated, 30-day average (basis: compressor kWh meter divided by flowmeter-integrated evaporation). Availability was 96% over twelve months on this clean duty, against 89% on a scaling geothermal duty of similar size where monthly washdowns consumed the difference — the same gap the scenario logic above predicts.

FAQ: Falling Film Evaporator Sizing and Limits

What minimum wetting rate should I design to?

Design to 0.1-0.3 m3/h per meter of tube perimeter (0.13-0.4 gpm/ft) for water-like, non-foaming liquors, and roughly double that for viscous or foaming feeds. The value is liquor-specific, so confirm it on a short-tube rig before fixing tube count. Operating below the minimum wetting rate produces dry patches, wall fouling, and a capacity loss that no control loop can recover.

How does falling film boiling differ from forced circulation?

In a falling film evaporator, liquor boils as a thin film on the inside of vertical tubes, with small holdup and 8-25 K of temperature difference per effect. In a forced-circulation evaporator, a pump moves liquor through the tubes below boiling, and flashing happens in the separator instead. That keeps crystals away from the tube wall, at the cost of pump power and lower heat-transfer coefficients.

What temperature difference per effect is typical?

Clean falling film effects run at 8-25 K of useful temperature difference; below about 8 K the extra heat-transfer area costs more than the steam it saves, and above roughly 25 K the film risks dryout and fouling. Concentrated sulfate or chloride liquors add 1-8 K of boiling point elevation, which must be subtracted before assigning area. Confirm the split against your liquor’s own boiling point rise curve.

Can a falling film evaporator crystallize salts?

Not at the tubes. Once liquor reaches saturation, wall boiling nucleates crystals on the hot surface and salting follows within days. The standard arrangement pairs the falling film effect with a separate forced-circulation or Oslo-type crystallizer where supersaturation is created by flashing, not by wall heat transfer. Falling film handles pre-concentration reliably up to about 90-95% of saturation, then hands over.

What is the viscosity limit for falling film operation?

Practical operation ends near 50-100 cP (mPa·s), depending on wetting rate and temperature difference. Above that, the film thickens, the wetting rate needed to keep the wall covered rises sharply, and the heat-transfer coefficient falls by half or worse. High-solids concentrates and late-stage brine concentration duties belong in forced circulation or scraped-film equipment instead. Always verify the limit at final concentration, not at feed concentration.

Does MVR work with falling film boiling?

Yes — mechanical vapor recompression on a falling film effect is the standard low-energy arrangement for clean liquors, because the small temperature lift matches a compressor’s economic range. Expect 25-45 kWh per tonne of water evaporated, assuming a 6-12 K lift and clean surfaces. Salting or heavily scaling feeds erode the advantage, since fouling forces frequent washdowns and pushes the compressor lift, and its power draw, upward.

How is the distribution plate designed?

The distributor must deliver equal flow to every tube within about 5-10%, usually through calibrated orifices or weir slots on a plate machined flat and installed level. Design checks include the feed turndown range, solids and viscosity excursions, and a full-wetting flush connection for shutdowns. A plate that is level in the shop but installed out of level by a few millimeters starves the uphill tubes permanently.

Conclusion

A falling film evaporator is the correct selection when the liquor is clean, the viscosity stays below about 50 cP, the duty stops short of saturation, and the wetting rate can hold 0.1-0.3 m3/h per meter of tube perimeter across the full turndown range — and the wrong one the moment any of those four conditions breaks. Size it from the wetting rate upward, not from a catalog area backward; subtract boiling point elevation before trusting the evaporator LMTD; size the vapor body from vapor volume at the coldest effect; and let a saturation line, not a preference, decide where forced circulation takes over. Worked honestly, that sequence turns a falling film evaporator into the cheapest tonne of water evaporated on the site. If your duty fits the scenarios above, request a quote with your feed analysis, flows, and target concentration, and our engineers will return a sized proposal with the arithmetic shown — or a candid recommendation that a forced-circulation stage serves you better. Either answer saves a project.

References

• Perry’s Chemical Engineers’ Handbook, 9th ed., McGraw-Hill (evaporation and vapor-recompression equipment sections).
• SME Mineral Processing and Extractive Metallurgy Handbook, Society for Mining, Metallurgy and Exploration, Englewood, CO.
• USGS, Mineral Commodity Summaries — Sodium Sulfate (annual edition).
• CSIRO, publications on brine concentration, saline water management, and salt recovery from mining waters.
• ALTA Metallurgical Services proceedings (ALTA Uranium/Rare Earth and salt-processing sessions), Perth.
• MetSoc, Annual Conference of Metallurgists proceedings (hydrometallurgy sessions).
• Hydrometallurgy journal, Elsevier.

Changelog

• 2026-11-05: Initial publication.