MVR Evaporator Selection: A Process Engineer’s Sizing Guide

Last updated: 2026-10-22

MVR evaporator selection — the sizing and choice of a mechanical vapor recompression (MVR) evaporator — comes down to matching three numbers: the evaporation duty in tonnes of water per hour, the boiling-point elevation of the liquor, and the compression ratio the compressor can deliver. Get this match right and the plant evaporates water for roughly 15–40 kWh of electricity per tonne of water, compared with 25–40 t of live steam per tonne for a single-effect unit.

This guide is one module in our evaporation and crystallization hub, where the full equipment family — from falling film to forced circulation and crystallizers — is mapped against hydrometallurgical duties.

Key Takeaways

• MVR cuts evaporation energy to 15–40 kWh per tonne of water evaporated, versus 25–40 t of steam for single-effect duty.
• Boiling-point elevation, not flow rate, sets the compression ratio and therefore the compressor type.
• Falling film calandrias suit clean, low-viscosity liquors; forced circulation suits scaling and saturating brines.
• Crystallizing duties such as sodium sulfate or lithium sulfate recovery need surge volume and anti-scaling design.
• Pilot testing on your actual liquor de-risks MVR evaporator sizing before capital is committed.

What Is Mechanical Vapor Recompression and How Does It Work?

Mechanical vapor recompression (MVR) is a heat-pump technique that recompresses the vapor boiled off a liquor so it can be reused as the heating steam in the same evaporator, over and over. A mechanical vapor recompression evaporator is therefore a closed energy loop: the only significant external energy input is electrical power driving the compressor, plus a modest start-up steam supply that is shut off once the loop reaches steady state.

how MVR vs Multi-Effect Evaporation is decided

The cycle works like this. Liquor boils at, say, 90 °C inside the calandria. The vapor leaving the separator is close to saturation at that temperature, but it carries far more latent heat than the boiling liquor needs at a slightly higher temperature. The MVR compressor raises the vapor pressure from roughly 70 kPa(abs) to perhaps 110–160 kPa(abs), which lifts the saturation temperature by 6–15 K depending on the duty. That superheated vapor then condenses on the outside of the heat-transfer tubes, and its latent heat boils more liquor. The condensate leaves as hot distilled water; the concentrate leaves from the bottom of the separator.

The figure that matters to an energy engineer is the coefficient of performance (COP), defined here as the mass of water evaporated per unit of compressor electrical energy expressed in steam equivalent. Practical MVR plants on water-like liquors achieve an evaporative economy of 15–40 kWh per tonne of water, which is thermodynamically equivalent to evaporating 15–30 t of water per tonne of equivalent steam. Because the energy carrier is electricity, MVR pairs naturally with sites that have low-cost power, waste heat from a smelter, or on-site renewables — a common situation in hydrometallurgical plants that also run electrowinning.

For a deeper physical walkthrough of the vapor loop, see our article on MVR working principle, and for the equipment-level options see the MVR evaporator product page.

Where MVR Evaporator Selection Fits in an Evaporation Plant Design Project

MVR evaporator selection is never done in isolation; it sits inside a wider evaporation plant design sequence that starts long before anyone quotes a compressor. In practice the sequence looks like this:

1. Characterize the feed: flow rate, dissolved solids, boiling-point elevation, scaling species, and corrosivity.
2. Define the target: concentrate to a fixed density, or reach full crystallization for zero liquid discharge (ZLD).
3. Screen evaporator types against the liquor chemistry — falling film, forced circulation, or a brine concentrator.
4. Set the energy frame: is steam cheap and available, or is electricity the more attractive carrier?
5. Only then size the MVR loop: compression ratio, heat-transfer area, and compressor type.

If steam is genuinely cheap on your site — for example, a smelter already raising surplus steam that is otherwise let down to atmosphere — a multiple-effect evaporator may still win the life-cycle comparison. That trade is examined side by side in our MVR vs multiple-effect comparison. When electricity is cheap or steam infrastructure is absent, MVR almost always dominates, which is why brine concentration MVR schemes have become the default for new lithium and nickel salt plants.

When Does MVR Make Sense for a Hydrometallurgical Liquor?

The honest answer is: MVR makes sense when the boiling-point elevation (BPE) of the liquor stays moderate, because BPE is the thermodynamic tax you pay on every tonne of vapor you recompress. BPE is the difference between the boiling temperature of your liquor and that of pure water at the same pressure. Pure water at 70 kPa(abs) boils at about 90 °C; a lithium sulfate mother liquor at 20 wt% Li2SO4 might boil 3–5 K hotter, and a saturated sodium sulfate stream can show 6–8 K, dominated by dissolved salt rather than by any suspended solids.

Rule-of-thumb guidance, based on published operating data for hydrometallurgical and ZLD plants, looks like this:

• BPE below roughly 8 K across the whole concentration range: single-stage MVR is straightforward.
• BPE between 8 K and about 15 K: MVR still works, but the compressor must deliver a higher compression ratio, and the effective COP drops; a two-effect MVR arrangement (two vessels in series sharing one compressor) may become attractive.
• BPE above roughly 15 K, or liquors that foam violently or boil with severe entrainment: MVR loses its edge, and a brine concentrator with its own vapor-compression loop, or simple steam-driven evaporation, deserves a hard look.

A second gate is saturation behaviour. A falling film evaporator is a vertical-shell evaporator that distributes liquor as a thin film down the inside of heated tubes, giving high heat-transfer coefficients at low liquid inventory — ideal for clean, non-scaling liquors. A forced-circulation evaporator, by contrast, is a pump-driven loop evaporator that keeps crystals suspended and sweeps heated surfaces at high velocity, which is why it is the standard vessel for liquors that will crystallize in the evaporator itself. If your duty is to concentrate a lithium sulfate raffinate ahead of a separate crystallizer, a falling film MVR unit is usually sufficient. If you intend to precipitate sodium sulfate cake directly from the MVR vessel, forced circulation with an external crystallizer volume is the safer selection — this interface is discussed in our crystallizer design article.

A third gate is corrosivity. Chloride-bearing liquors at elevated temperature push you toward duplex or super-austenitic stainless steels, titanium, or lined vessels, and these choices materially change the capital figure discussed later.

Key Design Parameters for MVR Evaporator Sizing

MVR evaporator sizing is an iterative loop between the heat balance and the compressor curve, but every iteration starts from the same small set of parameters. The table below is the checklist we apply at the opening of any MVR evaporator selection study; it lists the parameters a process engineer should fix first, with typical ranges observed in industrial brine and hydrometallurgical service. All values are indicative and must be confirmed for your liquor.

(Illustrative ranges compiled from published vendor literature and standard references such as Perry’s Chemical Engineers’ Handbook; they are not guarantees for any specific liquor.)

Two couplings in this table deserve emphasis. First, the heat-transfer area scales roughly inversely with the effective temperature difference, while compressor power scales roughly proportionally with it — so an aggressive designer who asks for a small vessel pays for it at the meter for the life of the plant. Second, the BPE that matters is the BPE at final concentration, not at feed concentration; a liquor that starts at 3 K and finishes at 12 K will have a compressor sized by the end point, and the earlier stages of the batch or the early modules of a multi-stage train will run with surplus headroom.

How Do You Select the MVR Compressor?

MVR compressor selection is the single most consequential equipment decision in the whole plant — in fact, compressor choice is where MVR evaporator selection succeeds or fails, because the compressor is typically 40–60% of the installed equipment cost and essentially all of the rotating risk. Four machine families dominate industrial practice.

Centrifugal fans (single-stage, high volume). These are the workhorses for large duties on low-BPE liquors — think 20–100 t/h of water from dilute feeds. They achieve compression ratios up to about 1.3, run at modest tip speeds, tolerate some entrained droplets, and are mechanically simple with few wearing parts. Most municipal and industrial ZLD brine concentrators use machines of this family.

Roots-type (lobed) blowers. Positive displacement with moderate pressure rise per stage, tolerant of wet vapor, and simple to control by speed. They are common on small and mid-size duties where capital is tight, but they are noisier and less energy-efficient than centrifugal machines at large volumes.

Screw compressors. Oil-free screw machines handle compression ratios up to roughly 1.8 in a single stage and cope well with high molecular weight vapors and some liquid carryover. For liquors with BPE in the 8–15 K band, screw compressors are frequently the correct MVR compressor selection because they deliver the temperature lift in one machine without interstage complications.

Multi-stage centrifugal compressors. When the duty demands a temperature lift beyond what one impeller can provide — high-salinity brines, deep concentration swings — multi-stage machines with intercooling or two vessels in series are specified. They carry the highest capital and maintenance complexity, so the extra lift must be justified by the heat balance.

Beyond machine type, three verification steps protect the project. First, check the compressor curve against the whole operating envelope, from start-up dilution to end-of-run concentration, not just the nameplate point. Second, confirm materials and seals against the vapor chemistry — acid mists, ammonia slip, or chloride carryover have ended more than one compressor life early. Third, verify the turndown path: a plant that must operate at 40% feed rate needs a compressor with inlet guide vanes or speed control that stays out of surge across that range. Vendor sizing for the exact vapor properties should always be counter-checked by the owner’s engineer; our MVR working principle article includes the thermodynamic checklist we use internally.

MVR Evaporator Selection Decision Rules

When the studies pile up, a disciplined decision sequence keeps the project honest. The following rules reflect how experienced engineers approach MVR evaporator selection on hydrometallurgical duties:

1. Fix the liquor data before the layout. No MVR evaporator selection should start without a verified analysis of the feed: full cation/anion assay, boiling-point elevation curve versus concentration measured in the lab, density–viscosity data, and a scaling-species assessment (CaSO4, SiO2, fluoride). A one-week lab program here is the cheapest risk reduction on the project.
2. Choose the evaporation temperature first, the compressor second. Higher evaporation temperature (90–100 °C where the product allows) gives a better heat-transfer coefficient and smaller area; product thermal stability and fouling tendency set the ceiling. The compressor is a consequence of this choice, not a starting point.
3. Match the calandria to the scaling risk, not to the catalogue. Clean, non-scaling liquors justify falling film tubes and their excellent heat transfer; any liquor within a factor of two of CaSO4 or SiO2 saturation should push the design toward forced circulation or a seeded, slurry-tolerant design.
4. Size the compressor at the worst corner of the operating window. The end-of-run BPE, minimum available suction pressure, and maximum fouled heat-transfer surface define the true design point; the nameplate point is a marketing artifact.
5. Decide crystallization inside or outside the MVR loop early. If solids will form in the evaporator, allow the extra volume, circulation pumping, and crystal-handling equipment from day one — retrofitting slurry capability into a clarifying evaporator is rarely successful.
6. Verify materials against the concentrate, not the feed. The most corrosive composition the vessel ever sees is the final concentrate at operating temperature; specify alloys against that, with a corrosion allowance, and have the coupon tests run in real liquor.
7. Insist on a pilot run for any new liquor chemistry. One to two weeks of piloting on actual plant liquor validates BPE, foaming, fouling rate, and crystal behaviour — the four inputs no desktop calculation can be trusted to predict.

Rule 7 deserves its own paragraph. Bench-scale boiling tests are useful for BPE and chemistry, but fouling rate and foaming are hydrodynamic phenomena that only appear at industrial velocity and vapor loading. A skid-mounted pilot evaporator running 50–200 L/h of your real liquor will surface these behaviours in days. Lexin-tech offers exactly this service — you can talk to us about pilot testing before committing to a full-scale design.

Which Failure Modes Should You Design Against?

Most MVR plant disappointments trace back to a short list of failure modes, each of which has a known countermeasure. A rigorous MVR evaporator selection process treats this as a pre-mortem checklist during design reviews:

1. Calcium sulfate scaling of heating surfaces. CaSO4 has inverse solubility and plateaus on hot walls rapidly once supersaturation is crossed. Countermeasures: keep the wall temperature difference below the scaling threshold (typically ≤10 K for suspect liquors), seed the liquor, or move to forced circulation; monitor via heat-transfer coefficient trend and clean on schedule. Our evaporator scaling control article details the chemical and operational options.
2. Silica fouling in the concentrate circuit. SiO2 polymerizes unpredictably above roughly 100–150 mg/L depending on pH and temperature, forming glassy deposits that resist chemical cleaning. Countermeasure: a desilication pretreatment step, or a deliberate silica bleed, before the MVR loop.
3. Compressor surge during turndown. Rapid feed reduction moves the machine toward surge, which can damage impellers within seconds. Countermeasures: inlet guide vanes, speed control with an anti-surge recycle line, and an interlock that trips to safe recirculation.
4. Foaming and entrainment carryover. Surfactant traces or high-salt viscosities destabilize the vapor space; entrained droplets contaminate condensate and, worse, reach the compressor. Countermeasures: generous vapor-space design, demisters sized for the actual surface tension, and online condensate conductivity monitoring with automatic divert.
5. Loss of vacuum or non-condensable accumulation. Air ingress from flanges and seals raises the condensing pressure, cuts heat transfer, and forces the compressor up its curve. Countermeasures: a properly sized vacuum system with continuous non-condensable venting from the condenser side.
6. Crystal carryover into the circulation loop. In crystallizing service, undersized settlers or apexes let solids erode pump internals and pack heat-transfer channels. Countermeasures: correct fines-removal loop, cyclone or screen classification, and velocity verification at minimum flow.
7. Corrosion under-deposit in chloride liquors. Local chemistry under scale deposits can be far more aggressive than the bulk liquor. Countermeasures: alloy selection with crevice data, disciplined cleaning cycles, and periodic inspection ports.

None of these modes is exotic; all of them are documented across the ZLD and hydrometallurgy literature. What separates a reliable plant from a troubled one is whether the designer assumed they would happen.

What Does an MVR Evaporation Plant Cost to Build and Operate?

Cost questions dominate late-stage discussions, so it is worth separating the fixed and variable sides honestly. The figures below are illustrative, drawn from the range of published benchmarks for stainless-steel equipment in similar service; they are not quotations and will vary with size, materials, and site conditions.

(Illustrative figures for planning discussion only; request a firm proposal for project numbers.)

The comparison that usually decides the debate is MVR against multiple-effect evaporation (MEE). An MEE train consumes roughly 0.3–0.5 t of steam per tonne of water evaporated per effect-times-two, so a four-effect unit needs about 0.3 t of steam per tonne of water; MVR needs no steam at all in continuous service but consumes electricity. The crossover therefore depends entirely on the steam-to-electricity price ratio at your site, plus the capital difference. Our MVR vs multiple-effect comparison walks through the break-even arithmetic, and our evaporator energy consumption article benchmarks the per-tonne numbers in more detail. For sites targeting zero liquid discharge, the evaporation stage sits upstream of crystallization and drying, and the energy carrier chosen here propagates through the whole chain.

Worked Sizing Example (Illustrative)

The following example shows the arithmetic chain a process engineer runs during MVR evaporator sizing. All numbers are invented for illustration and do not describe any real project or customer.

Task: concentrate 20 m³/h of lithium sulfate raffinate from 6 wt% to 24 wt% Li2SO4 ahead of a crystallizer, evaporation temperature 92 °C, plant electricity at 0.07 USD/kWh.

1. Water balance. Feed solids: 20,000 kg/h × 0.06 = 1,200 kg/h Li2SO4. At 24 wt%, concentrate flow = 1,200 / 0.24 = 5,000 kg/h. Water evaporated = 20,000 − 5,000 = 15,000 kg/h = 15 t/h.
2. Boiling-point elevation. From lab data for this liquor (illustrative), BPE at 24 wt% is 4.5 K; at 6 wt% it is 1.5 K. Design to the end point.
3. Heat balance at ΔT = 7 K. Assume overall heat-transfer coefficient U = 1.6 kW/m²·K for falling film at this viscosity (illustrative, from Perry’s-range correlations). Latent heat of water at 92 °C ≈ 2,278 kJ/kg. Heat duty = 15,000 kg/h × 2,278 kJ/kg ÷ 3,600 = 9,492 kW. Area = Q / (U × ΔT) = 9,492 / (1.6 × 7) ≈ 848 m².
4. Compression ratio. Suction at 75 kPa(abs) (92 °C saturation minus BPE correction, illustrative); discharge needs saturation at 92 + 7 = 99 °C plus ~1 K for pressure drop, i.e. about 103 °C, which is roughly 113 kPa(abs). Compression ratio = 113 / 75 ≈ 1.51.
5. Compressor power (illustrative). Isentropic compression of steam with overall efficiency 70% gives approximately 25 kWh per tonne of water evaporated at this ratio (consistent with vendor curves for screw machines). Compressor power = 15 t/h × 25 kWh/t ≈ 375 kW installed draw.
6. Energy check and economy. Equivalent single-effect steam use would be about 15 t/h of steam; MVR uses 375 kW ≈ 1.35 GJ/h versus roughly 34 GJ/h of steam latent heat — a better than 20-fold reduction in primary energy, in line with the 15–40 kWh/t benchmark quoted earlier.
7. Operating cost. 375 kW × 0.07 USD/kWh × 8,000 h/y ≈ 210,000 USD/y for evaporation energy. A single-effect alternative at 15 t/h steam would face steam at even a modest 25 USD/t: 15 × 8,000 × 25 = 3,000,000 USD/y. This 1.4-million-dollar annual gap is why lithium carbonate and lithium sulfate plants have converged on MVR for their concentration duties.

Note where the sensitivities sit: a 2 K increase in achievable ΔT (better U, or cleaner tubes) removes about 200 m² of area, while a 2 K increase in BPE pushes the compression ratio and specific power upward for the life of the plant. That asymmetry is the essence of MVR evaporator sizing — spend effort on liquor data and fouling control before you spend money on steel.

Frequently Asked Questions

What evaporation duties suit MVR best?
MVR performs best on large, continuous duties with moderate boiling-point elevation, typically below 8–10 K across the concentration range, and with low scaling risk or good fouling control. Clean filtrates, reverse-osmosis rejects, and metal salt mother liquors ahead of crystallization are classic fits. Duties with severe foaming, extreme BPE, or aggressive corrosives need special design attention or a different evaporator family altogether.

Evaporator Crystallizer Systems for Mining Brines

How much electricity does an MVR evaporator actually use?
Installed plants typically report 15–40 kWh of compressor electricity per tonne of water evaporated, with the low end on dilute, low-BPE liquors and the high end on concentrated, high-BPE brines with higher compression ratios. Converted to primary energy, that is equivalent to evaporating 15–30 t of water per tonne of steam, which is why MVR is considered the benchmark for evaporation energy efficiency in continuous service.

Can MVR reach full crystallization, or only concentration?
Both are done industrially, but crystallizing duty changes the design substantially. If solids form inside the evaporator, the vessel needs slurry-tolerant circulation — usually forced circulation — plus fines management and crystal washing downstream. Many plants instead concentrate to just below saturation in an MVR unit and send the hot concentrate to a separate crystallizer, which keeps the evaporator clean and simplifies crystal quality control.

What happens to non-condensable gases in an MVR loop?
Gases such as air, carbon dioxide, or ammonia slip do not condense and accumulate in the steam side, raising condensing pressure and choking heat transfer. Every MVR plant therefore includes a vent system — typically a small steam-jet or liquid-ring pump — that continuously bleeds non-condensables from the condensing side. The energy penalty is small, but forgetting the vent is a classic commissioning problem.

How long does a typical MVR compressor run between overhauls?
Depending on machine type and vapor cleanliness, overhaul intervals of 20,000–40,000 operating hours are common for centrifugal fans and screw machines, with Roots blowers somewhat less. Machines handling droplet-laden or corrosive vapor sit at the low end. Planning maintenance around scheduled overhauls, with condition monitoring on vibration and discharge temperature, keeps availability above 95% in well-run plants.

Is MVR suitable for intermittent or batch operation?
MVR rewards continuity. Each start-up consumes steam to establish the vapor loop, and cycling increases thermal stress and fouling. Plants operating fewer than roughly 4,000 hours per year, or with strongly variable feed, should evaluate whether a simpler steam-driven evaporator, or a mixed scheme with steam assist during ramps, gives a better life-cycle cost than a pure MVR loop.

How does brine concentration MVR differ from a brine concentrator?
The terms overlap, but convention distinguishes them. “Brine concentration MVR” usually denotes a single-stage mechanical vapor recompression evaporator concentrating a brine partway, often ahead of crystallization. A brine concentrator is a specialized high-recovery vertical-tube or spray-film MVR machine that pushes brines to very high dissolved-solids content, sometimes 20–25 wt%, with aggressive anti-scaling features. The right choice depends on target recovery and scaling species.

Conclusion

MVR evaporator selection is ultimately an exercise in respecting three numbers — duty, boiling-point elevation, and compression ratio — and letting them, rather than habit or catalogue enthusiasm, dictate the equipment. A liquor with modest BPE and honest fouling data can justify a straightforward falling film MVR plant that evaporates water for a few tens of kilowatt-hours per tonne; a scaling, high-BPE brine needs forced circulation, a more capable compressor, and a pilot test before steel is cut. Work through the decision rules in order, verify the compressor at the worst operating corner, and validate fouling and foaming behaviour on real liquor. When you are ready to move from desktop study to evidence, talk to us about pilot testing or browse the wider resource library for the supporting articles in this cluster. A well-chosen MVR plant is one of the few investments in a hydrometallurgical flowsheet that pays for itself every single operating hour.

References

1. Green, D. W., and Southard, M. Z. (eds.), Perry’s Chemical Engineers’ Handbook, 9th ed., McGraw-Hill — evaporation, vapor recompression, and heat-transfer design chapters.
2. McCabe, W. L., Smith, J. C., and Harriott, P., Unit Operations of Chemical Engineering, 7th ed., McGraw-Hill — evaporator energy balances and multiple-effect design fundamentals.
3. Mersmann, A. (ed.), Crystallization Technology Handbook, Marcel Dekker — crystallizer selection and slurry-handling design guidance.
4. Hydrometallurgy (journal, Elsevier) — peer-reviewed literature on brine treatment, salt recovery, and evaporation-crystallization in metal recovery flowsheets.
5. U.S. Geological Survey, Mineral Commodity Summaries (annual edition) — lithium and sulfate salt production and market context.
6. Industrial vendor application literature on mechanical vapor recompression evaporators and brine concentrators (GEA, Veolia, and similar equipment suppliers, general catalogues) — compressor types, economy benchmarks, and ZLD case descriptions.

Changelog

2026-10-22 — First published.

Related reading

• Crystallization: selection rules