Mechanical Vapor Recompression: Working Principle & Energy

Last updated: 2026-10-25

Mechanical vapor recompression (MVR) typically reduces evaporator energy demand to 15-30 kWh per tonne of water evaporated, roughly 20-40 times less than the 600+ kWh/t steam-equivalent consumed by a conventional multi-effect train. Instead of dumping clean water vapor from the boiling liquor, an MVR system compresses that vapor a few degrees hotter and returns it as the heating steam. This page explains the working principle, compressor options, energy math, and where MVR wins in hydrometallurgical duty — part of our broader evaporation and crystallization coverage.

Key Takeaways

• MVR re-uses vapor as its own heating steam, cutting specific energy to 15-30 kWh/t water.
• Electricity replaces steam, so economics hinge on local power versus fuel prices.
• Boiling point elevation (BPE) sets the compressor lift and therefore the feasible ceiling.
• Fans, twin-screw and centrifugal compressors cover different flow and ΔT windows.
• Scaling, incondensables and foaming are the main field causes of performance drift.
• Pilot testing with real liquor de-risks BPE, viscosity and foaming before detailed design.

What Is Mechanical Vapor Recompression?

Mechanical vapor recompression is an energy-recovery technique that recompresses low-pressure vapor evaporated from a boiling liquor and returns it to the evaporator’s heating side as the driving steam. The system is, in effect, a heat pump applied to an evaporator: it takes vapor at saturation pressure, raises its pressure and saturation temperature by a few kelvin, and condenses it across the heat transfer surface to boil more feed. The compressor supplies only the work needed to bridge that small temperature lift, which is why the electrical input is a small fraction of the latent heat being recycled.

A conventional evaporator discharges vapor to a condenser, throwing away roughly 2,250-2,300 kJ per kg of water as low-grade heat, or consumes live steam in a multi-effect (MED) cascade to reuse it a handful of times. An MVR evaporator reuses that same vapor continuously — the evaporator body effectively becomes a closed thermodynamic loop with a compressor in it. One pass of steam economy is replaced by an essentially unlimited recirculation, bounded not by effect count but by compressor efficiency and heat transfer temperature differences.

The concept is not new. Mechanical recompression was patented in the 19th century in Europe and installed on brine and sugar duty in the early 20th century, but adoption accelerated from the 1970s onward as oil prices rose and industrial centrifugal fans and screw compressors matured. Today MVR is the default choice for bulk evaporation of clean or moderately contaminated liquors — dairy, ethanol stillage, pulp mill condensates, and increasingly lithium, nickel and sodium sulfate streams — wherever steam is expensive and the boiling point rise is modest.

For hydrometallurgy specifically, the technology matters because evaporation and crystallization are among the most energy-intensive steps in a flowsheet. A nickel sulfate crystallization circuit or a sodium sulfate recovery loop can be dominated by evaporative duty; an MVR retro-fit is frequently the single largest energy-saving lever available. The trade-offs — compressor type, BPE tolerance, scaling behavior — are covered in our dedicated MVR evaporator guide, and in the parameter table and calculation below.

Key Operating Parameters

Two terms frame everything that follows. Boiling point elevation (BPE) is a thermodynamic property that raises a solution’s boiling temperature above that of pure water at the same pressure, and it widens the temperature lift the compressor must deliver. Compressor surge is an aerodynamic instability that occurs when vapor flow through the machine falls below the stable operating line, producing oscillation, vibration and potential mechanical damage. Six parameters define whether an MVR evaporator will deliver its promised energy figures in a given service. Getting any one of them wrong at the flowsheet stage is the usual root cause of an underperforming plant.

MVR Evaporator Selection

Evaporating temperature and vapor volume. The compressor handles enormous volumetric flows because water vapor at, say, 70 °C has a specific volume near 5.5 m³/kg. Raising the evaporating temperature shrinks the machine dramatically — doubling absolute pressure roughly halves the volume flow per tonne evaporated. This is why falling film evaporators on MVR duty often run at 85-95 °C rather than the deep vacuum common in MED plants, provided the liquor tolerates the temperature.

ΔT and heat transfer area. The useful ΔT across the calandria is the compressed saturation temperature minus the boiling liquor temperature. Because MVR recycles its own vapor, generous area bought with cheap heat exchange surface directly buys lower compressor power — the classic optimization is trading plates or tubes against compressor kilowatts. Falling film geometry is preferred for viscous or moderately scaling liquors because it gives high heat transfer coefficients at small ΔT with short residence times.

Boiling point elevation. BPE is the difference between the liquor boiling temperature and pure-water boiling temperature at the same pressure. Pure water recovery has BPE near zero; a saturated sodium sulfate liquor shows roughly 3-5 K (typical); concentrated lithium brines and alkaline streams can exceed 8 K. BPE is “wasted” compression lift — it must be paid for on every kilogram of vapor, yet it does no evaporative work. When BPE plus process ΔT exceeds about 20-25 K, MVR economics typically collapse and MED with thermo-vapor recompression (TVR) becomes the better comparison; our MVR vs multiple-effect analysis walks through that crossover.

Compression ratio and machine selection. The ratio is set by absolute discharge pressure over suction pressure — small numbers, but achieving them on huge flows is a specialist task. Axial or high-flow centrifugal fans suit very large, low-ratio duties (waste water, condensate). Twin-screw compressors handle higher ratios, wet or fouling-prone vapors and wide turndown. Single-stage or two-stage centrifugal compressors cover mid-range duties in nickel sulfate and ZLD plants. Compressor efficiency — isentropic or polytropic — spans roughly 70-82% for well-designed machines, and it enters the energy calculation directly.

Specific energy and run length. Published vendor figures of 15-30 kWh/t for low-BPE duty are achievable, but a fouled calandria forces higher lift, and an unsteady compressor rides its surge line. Plants on scaling duty — high-sulfate, high-silica or calcium-bearing streams — need realistic clean-in-place (CIP) planning; our notes on evaporator scaling control cover antiscalant and chemical cleaning practice. Do not accept any specific-energy guarantee without a liquor-specific BPE measurement.

How Does an MVR Evaporator Work Step by Step?

An MVR evaporator looks complicated on a P&ID, but the working principle is a simple loop walked through in seven steps. The sequence below follows one kilogram of water from feed to condensate on a single-effect falling-film unit.

1. Feed preheat. Cold feed — for example a purified nickel sulfate solution from solvent extraction — enters a preheater bank and is heated by the outgoing condensate and, in many designs, a vent condenser on the non-condensable purge line. Recovering 15-20 K of sensible heat here is cheap insurance against parasitic steam trimming. Feed is brought to within a few kelvin of its boiling point before it reaches the calandria distribution system.

2. Vapor separation. Preheated feed flows down the inside of the falling film tubes where it partially evaporates, then the two-phase stream enters the vapor body. The separator — often a tangential-entry vessel or a plate-type demister — disengages liquid from vapor. Liquid recirculates or passes forward to the next effect or a crystallizer; the vapor, at evaporator pressure and saturation temperature, is the raw material for the whole MVR concept. Foaming liquors demand careful demister sizing, because droplet carryover salts up the compressor.

3. Compression. The vapor is drawn into the compressor — a centrifugal fan, twin-screw or centrifugal impeller machine depending on the duty — and raised by typically 5-15 K of saturation temperature equivalent. The work input is small relative to the latent heat carried, which is the entire economic argument. Compressor speed or inlet guide vanes modulate flow to follow the evaporative demand, keeping the machine away from surge at turndown.

4. Condensation on the other side. Superheated discharge vapor — usually a few kelvin of superheat above saturation — flows to the heating side of the calandria (the shell side of the falling film unit). Desuperheating is recovered as useful heat, then the vapor condenses on the tubes at the elevated saturation temperature, releasing its latent heat to boil more liquor on the other side of the same metal wall. The loop is now closed: yesterday’s reject heat is today’s heating steam.

5. Condensate removal. The condensate — essentially distilled water — drains from the shell side through level-controlled traps. In a hydromet plant this distillate is often a product stream: process water for washing, or feed to an RO polish. Condensate is flashed back into the loop via the preheater in step 1, so its sensible heat is recovered rather than sewered.

6. Non-condensables purging. Air leaks, dissolved gases liberated from the feed, and decomposition products accumulate in the loop and blanket the heat transfer surface. A small purge stream with a vent condenser removes them continuously. Purge losses are real: every kilogram of incondensable gas drags water vapor with it, and excessive purge air is one of the most common hidden energy sinks we find in audits of underperforming units.

7. Control loop. Product density (or boiling point rise, or both) controls residence and draw-off; evaporator level controls feed; and the compressor capacity control holds the balance between boil-off and heat transfer. A well-tuned loop holds constant ΔT so the compressor never chases the calandria into fouling-induced high lift. Instrumentation worth insisting on: suction and discharge pressure-temperature pairs on the compressor, liquor-side BPE indication, and differential pressure across the demister. For liquor-specific screening before committing capital, a bench or pilot campaign such as our pilot testing service validates steps 1-3 with the actual solution rather than water.

Why Does MVR Performance Degrade?

Field data from operating MVR plants show a predictable set of failure modes. Each one below is stated as symptom, then cause, then remedy — the order a commissioning engineer works through them.

1. Rising BPE and apparent capacity loss. Symptom: evaporation rate falls while the compressor runs hotter and draws more power; product density drifts upward. Cause: dissolved solids concentration has crept past design, raising boiling point elevation and viscosity, so the same compressor lift delivers less useful ΔT. This is common in sodium sulfate circuits where crystallization control has slipped and the liquor has drifted supersaturated. Remedy: verify BPE against the design curve, restore correct discharge density control, and check whether nucleation/seeding in the crystallizer section is doing its job.

2. Fouled heat transfer plates or tubes. Symptom: falling heat transfer coefficient over weeks, ΔT widening at constant duty, eventually a hard capacity limit. Cause: calcium sulfate, silica or crystal scale on the liquor side; occasionally organic films. Remedy: chemical CIP with an appropriate acid or chelant recipe, mechanical cleaning where chemistry fails, and prevention through antiscalant dosing, seed control and upstream softening — the practices detailed in our scaling-control resource. Design the exchanger for cleanable geometry from day one; a plate pack you cannot open is a liability on scaling duty.

3. Compressor surge or vibration. Symptom: rumbling, oscillating discharge pressure, high vibration trips at reduced throughput. Cause: operating below the machine’s surge flow at turndown, or a suction side restriction from demister blinding by carryover. Remedy: check the antisurge recycle valve and capacity control logic, clean the demister, and re-verify the operating point against the vendor map. Twin-screw machines tolerate this mode better than high-speed centrifugal impellers, one reason screw units are favored on fouling-prone brines.

4. Incondensables accumulation. Symptom: gradual loss of top-end heat transfer, high shell-side pressure, steam-saturated purge. Cause: air ingress at seals, CO₂ or NH₃ released from feed, undersized or mis-operated purge. Remedy: leak-hunt the vacuum envelope, confirm purge flow against the design basis, and use the vent condenser to keep water loss minimal. A few percent of air in the heating steam can cut the condensing coefficient dramatically — this is the cheapest fault to fix and the most frequently ignored.

5. Foaming and vapor carryover. Symptom: salt-laden condensate, compressor fouling or imbalance within days of startup, demister ΔP alarms. Cause: surfactants, finely divided solids or organic content in feed stabilizing foam at the vapor-liquid interface. Remedy: antifoam dosing sized and injected correctly (overdosing fouls the demister in a different way), demister upgrade, or feed pretreatment. Carryover in an MVR loop is self-amplifying because the compressor is inside the loop — the salts deposit exactly where the machine is least tolerant of them.

None of these modes is exotic; all are visible in trend data weeks before they cost production. A monthly review of compressor power per tonne evaporated, BPE, and demister ΔP catches nearly every one early.

How Do You Calculate Mechanical Vapor Recompression Energy Demand?

The energy calculation for a mechanical vapor recompression evaporator reduces to dividing the latent heat recycled by the effective coefficient of performance (COP) of the compressor, then adding motor and gear losses. Worked example below — assumptions are stated so you can substitute your own.

how MVR vs Multi-Effect Evaporation is decided

Basis. Evaporate 10 t/h of water at an evaporating temperature of 90 °C from a liquor with BPE of 2 K. Latent heat of vaporization at 90 °C ≈ 2,283 kJ/kg. Assume a useful ΔT across the heat transfer surface of 6 K, so the compressor must lift the saturation temperature by ΔT + BPE = 8 K. Assume isentropic (compression) efficiency 75%, motor and gear efficiency 95%.

Step 1 — ideal compressor work. For small lifts, the isentropic work per kilogram of vapor is approximated by w = c_p · ΔT_s / η_s, where ΔT_s is the temperature rise of the vapor during compression. For a lift of 8 K saturation equivalent and superheat behavior of steam vapor, an isentropic enthalpy rise of roughly 18-20 kJ/kg is typical; take 19 kJ/kg. Actual work = 19 / 0.75 ≈ 25.3 kJ/kg.

Step 2 — electrical power at the motor. Mass flow of vapor = 10,000 kg/h. Compression power = 10,000 × 25.3 = 253,000 kJ/h = 70.3 kW (isentropic basis); with motor/gear losses at 95%: 70.3 / 0.95 ≈ 74 kW electrical.

Step 3 — specific energy. 74 kW ÷ 10 t/h = 7.4 kWh per tonne of water evaporated. Real installations on clean duty report 15-30 kWh/t (typical) because real lifts, part-load operation, purge losses, fouling allowance and auxiliary consumers (vacuum system, pumps, controls) all add on top of the thermodynamic minimum. Using the conservative 15-30 kWh/t band for planning is prudent; 7.4 kWh/t is the idealized floor for this duty.

Step 4 — comparison with MED steam energy. A quadruple-effect evaporator achieves a steam economy near 4.0, consuming about 0.25 t of steam per tonne evaporated. Steam latent heat ≈ 2,200 kJ/kg → 0.25 × 2,200 × 10,000 kg/h = 5.5 GJ/h thermal. Converting at fuel-to-steam efficiency of 90% and expressing as primary energy, that is roughly 600-640 kWh of fuel energy per tonne of water — 25-80× the MVR electrical demand depending on which end of the band you take. Even comparing on purchased-energy cost, MVR usually wins wherever electricity is not grossly expensive relative to fuel; the crossover logic and payback math are laid out in our evaporator energy consumption note.

Two cautions: first, COP of an MVR compressor is not a fixed number — it degrades with lift and with fouling, so quote specific energy at a stated ΔT and BPE; second, where the liquor’s BPE is high, rerun the numbers before believing any vendor claim, because lift enters the enthalpy rise directly.

What Do Installed Plants Show?

Published vendor figures are useful, but plant operating data tell the real story. The numbers below are illustrative and anonymized, drawn from the pattern of hydrometallurgical and industrial brine installations; treat them as representative orders of magnitude rather than guarantees.

First, an MVR evaporator concentrating lithium-bearing brine in a high-altitude South American operation (at roughly 4,000 m elevation, where reduced ambient pressure affects vacuum system design) reported specific energy in the low 20s of kWh/t — within the realistic band — after two rounds of demister and purge optimization in the first year. The second-year improvement came mostly from fixing air in-leakage, not from any thermodynamic change.

Second, a nickel sulfate refining project in East Asia feeding an MVR falling film unit ahead of a forced-circulation crystallizer achieved its design evaporation rate at approximately 25 kWh/t, but only after CIP frequency was doubled against the original schedule. The lesson from that installation: the compressor was never the problem — the fouling model in the design phase underestimated calcium and silica carryover from upstream filtration upsets.

Third, a sodium sulfate recovery plant in Inner Mongolia operating an MVR crystallization train showed the classic BPE signature: as liquor concentration cycled with crystallizer control swings, measured BPE moved by 1.5-2 K and specific energy followed it upward by roughly 10%. Tightening discharge density control recovered the loss without any hardware change — a control solution worth roughly the cost of one month of excess electricity.

Fourth, a ZLD brine concentration service in the Middle East running a large twin-screw compressor machine demonstrated the value of ratio headroom: the unit was originally selected for 12 K lift, and because the machine carried margin, a later switch to a higher-salinity feed (BPE up ~3 K) required no compressor change — only recalibrated controls and a revised antiscalant program. These four patterns — purge discipline, fouling realism, BPE-driven control, and ratio margin — recur across continents and are the practical bridge between the calculation above and a plant that hits its numbers.

Frequently Asked Questions

What is mechanical vapor recompression in simple terms?
It is a heat pump bolted onto an evaporator. The water vapor boiled off the liquor — normally waste heat — is compressed by a few kelvin of saturation temperature and used again as the heating steam on the other side of the heat exchanger. Because the latent heat is recycled continuously, the only significant energy input is the compressor’s electricity, typically 15-30 kWh per tonne of water evaporated on clean duty.

When does MVR beat a multi-effect evaporator?
MVR wins when steam is expensive relative to electricity, the liquor’s boiling point elevation is modest (roughly under 8-10 K), and the plant runs continuously. MED remains competitive where cheap waste steam or fuel exists, when BPE is very high, or for intermittent operation. As a planning heuristic, MVR saves its capital premium within one to four years in most power-price environments — but verify with site tariffs and a liquor-specific calculation.

How much electricity does an MVR evaporator use per tonne of water?
Thermodynamic minimums for low-BPE duty are below 10 kWh/t, while realistic operating plants report 15-30 kWh/t (typical) once fouling allowance, purge losses, part-load operation and auxiliary consumers are included. The exact figure depends on compression lift, isentropic efficiency and motor losses. Any vendor guarantee should be referenced to a stated ΔT, BPE and clean-versus-fouled heat transfer condition.

What boiling point elevation can MVR handle?
Commercial machines comfortably handle combined lift (ΔT plus BPE) in the 8-20 K range; twin-screw and multi-stage centrifugal compressors extend this somewhat. As BPE rises, compressor power rises roughly proportionally, so high-BPE liquors — concentrated caustic, some saturated brines — need a careful economic check against alternatives. Measure BPE on the actual liquor at final concentration; textbook values for pure salts often underpredict mixed-industrial streams.

Which compressor types are used in MVR systems?
Three families dominate. Centrifugal fans and axial machines handle very large vapor flows at low compression ratio (roughly 1.2-1.5). Single- or two-stage centrifugal compressors cover mid-range duties with high efficiency. Twin-screw compressors tolerate higher ratios, wet or fouling vapors and deep turndown, which makes them common on brine and crystallization service. Selection hinges on vapor volume flow, required lift, and tolerance to carryover.

Does MVR work for crystallizing liquors like sodium sulfate?
Yes — MVR-driven forced-circulation and osmosis-assisted crystallizers are standard in sodium sulfate and similar salt recovery. The crystallizer operates at controlled supersaturation with seeding or fines destruction, and the MVR loop supplies the evaporative duty. The main design difference from a plain evaporator is managing nucleation and crystal slurry density so the heat exchange surfaces stay clean and the BPE stays predictable.

How often does an MVR evaporator need cleaning?
Run length between clean-in-place cycles ranges from about a week on severe scaling duty (calcium, silica, high sulfate) to three months or more on clean organic streams. Trend the heat transfer coefficient monthly; schedule CIP on measured fouling, not on the calendar. Antiscalant dosing, feed softening and demister maintenance extend run length more cheaply than oversizing the compressor to muscle through fouling.

Can an existing multi-effect evaporator be retrofitted with MVR?
Often, yes. A common retrofit replaces the first-effect live steam with a compressor on the last-effect vapor, or converts the train to single-effect MVR with the existing vessels reused as surface or preheat. The feasibility screen covers compressor fit to available vapor flow, vessel pressure ratings, and whether the existing heat transfer area supports the smaller ΔT. Expect an energy reduction of 60-85% on steam-equivalent primary energy in favorable cases.

Conclusion

Mechanical vapor recompression converts evaporation from one of the largest steam consumers in a hydrometallurgical flowsheet into one of its most efficient operations, holding specific energy to 15-30 kWh per tonne of water evaporated against hundreds for steam-fired alternatives. The working principle is elegant, but plant performance is decided by less glamorous details: measured BPE on the real liquor, fouling behavior, purge discipline and compressor selection matched to the lift. Lexin-tech designs and supplies MVR evaporators, crystallizers and complete zero liquid discharge systems for nickel sulfate, sodium sulfate and complex brine duty, backed by pilot testing with your actual solution. If you are evaluating evaporation duty, send us the liquor analysis and flow basis — we will return a specific energy calculation and selection recommendation you can defend to your project team. Further reading is available in our resource library.

Evaporator Crystallizer Systems for Mining Brines: selection rules

References

1. Perry, R.H. & Green, D.W., Perry’s Chemical Engineers’ Handbook, 9th ed., McGraw-Hill — sections on evaporation, vapor recompression and evaporator economics.
2. Coulson, J.M. & Richardson, J.F., Chemical Engineering, Volume 2 (Particle Technology and Separation Processes), Butterworth-Heinemann — evaporator types, multiple-effect design, boiling point elevation.
3. Kern, D.Q., Process Heat Transfer, McGraw-Hill — fundamentals of condensing-side heat transfer and ΔT optimization in evaporators.
4. GEA / industry evaporation engineering literature — general treatment of mechanical vapor recompression, compressor types and specific energy bands for falling film evaporators.
5. Minton, P.E., Handbook of Evaporation Technology, Noyes Publications — evaporator selection, energy recovery and recompression practice.
6. Kucera, J., Reverse Osmosis: Industrial Applications and Processes, Wiley-Scrivener — context for brine concentration and ZLD integration upstream of evaporation.
7. U.S. Geological Survey, Mineral Commodity Summaries (lithium, nickel) — demand context for lithium brine and nickel sulfate evaporation circuits.
8. Desalination and Applied Thermal Engineering (journal literature generally) — peer-reviewed studies on vapor compression heat pumps, compressor efficiency and evaporator energy performance.

Changelog

• 2026-10-25 — first published.

Related reading

• Crystallization: selection rules