Take the steam bill for a concentration plant, divide it by the tonnes of water removed, and you have the only number that really matters when buying an evaporator. Everything else on the quotation is a detail by comparison.
Evaporation is a brute force operation. You are boiling water off a solution, and the latent heat of vaporisation does not negotiate. A plant that removes twenty tonnes of water an hour is buying a great deal of energy every hour it runs, for as long as it runs. Which is why the capital cost of the equipment, the number most procurement processes are built around, is frequently the smaller half of the decision.
One number, and what it means
Steam economy is kilograms of water evaporated per kilogram of steam supplied. It is the efficiency of the whole system in a single figure and it is the right basis for comparing two competing offers.
Boil water with steam in a single vessel and you get a steam economy slightly below one. You put in roughly a kilogram of steam to remove roughly a kilogram of water, with the shortfall going to heating the feed and to losses. That is the baseline, and it is expensive.
The insight the whole industry is built on is that the vapour boiled off in the first vessel is itself steam, at a lower pressure, still carrying almost all of its latent heat. If you can use it to boil something else, you get a second evaporation for no additional fuel. That is a multiple effect evaporator: a series of bodies at successively lower pressures, each heated by the vapour from the one before.
The arithmetic is close to linear. Two effects roughly double the steam economy, three roughly triple it. A five effect plant will land somewhere around four and a half kilograms of water per kilogram of steam, which is to say it uses about a fifth of the energy of a single effect doing the same job. We build multiple effect systems from two effects through six and beyond.
So why does anyone stop at five effects
Because each effect you add costs a vessel, and because a physical limit closes in from the other direction.
Every effect needs a temperature difference across its heating surface to drive heat transfer. The total temperature difference available to you is fixed at the top by the steam you can supply and at the bottom by the temperature of your cooling water in the condenser. Add effects and you divide that same total difference into more, smaller slices. Each slice gets thinner, and thinner slices need proportionally more heating surface, so the equipment gets larger and more expensive precisely as the energy saving per added effect gets smaller.
Boiling point elevation makes it worse. A concentrated solution boils above the temperature of pure water at the same pressure, and that elevation is subtracted from your available driving force in every single effect. On solutions with strong boiling point elevation, the useful number of effects is limited by chemistry rather than by budget.
The optimum is therefore a genuine calculation, not a preference, and it depends on the cost of steam at that specific site.
Mechanical vapour recompression changes the question
There is another way to reuse the vapour. Instead of passing it down a chain of effects, compress it. Raising the pressure of the vapour raises its condensing temperature, which means it can be fed back to heat the very same body it just came from.
A mechanical vapour recompression system does this with a compressor driven by an electric motor. Once it is running, live steam is needed only for start up. Energy consumption per tonne of water becomes very low, and in energy terms MVR outperforms almost any number of effects you would sensibly build.
The trade is threefold. You are converting a steam demand into an electrical demand, so the economics turn on the local ratio between electricity and fuel prices. You have introduced a large rotating machine into a plant that previously had none, with the maintenance and reliability implications that carries. And MVR works best across a modest temperature lift, which means it suits solutions with low boiling point elevation better than heavily concentrated ones.
Thermal vapour recompression is the simpler cousin. A steam ejector, with no moving parts, entrains part of the vapour using high pressure motive steam. It recovers less than MVR and costs a fraction as much, and it is frequently combined with a multiple effect train to lift the overall economy a useful step.
The body type follows the liquid
Steam economy is a system decision. The type of evaporator body is a product decision.
Falling film evaporators run a thin film down the inside of vertical tubes. They give short residence time, low temperature difference operation and gentle treatment, which is why they dominate dairy, juice and other heat sensitive duties. They demand even distribution to every tube, and a dry patch on a tube wall will foul immediately.
Rising film units use the vapour generated at the bottom to drag liquid upward as a film. They are simple and effective within their range.
Forced circulation evaporators pump the liquid through the heating element fast enough to suppress boiling inside the tubes, with flashing taking place in a separator instead. This is the answer for solutions that scale, crystallise or are highly viscous, because keeping the surface wetted and the velocity high is what stops deposition. It costs pumping power and it earns it back in run length.
Agitated thin film evaporators use rotating blades to spread a very thin film mechanically. For extremely viscous, fouling or heat sensitive products they can process material that no other type will handle.
Calandria designs, natural and forced circulation, remain standard across sugar and several bulk chemical duties. We build calandrias with diameters beyond 7,000 mm and evaporator shells past 8,000 mm, from full vacuum to 300 bar and above, in single pieces up to 500 MT, along with crystallisers and the full condenser range covering surface, vacuum, process and cryogenic service.
Scaling is the operating cost nobody quotes
An evaporator concentrates a solution, which means it deliberately drives the dissolved solids towards saturation. That is the process working as intended. It is also exactly the condition in which scale forms on a heated surface.
Scale is worse here than fouling elsewhere, because it grows precisely where the driving force is largest and it grows on the surface you are paying to maintain. Performance decays gradually, capacity slips, and eventually the plant is taken off line for cleaning. In a continuous operation, that shutdown is often the single largest real cost of the whole system.
The design responses are known. Keep velocity up, which is the forced circulation argument. Avoid local hot spots and excessive temperature difference across the surface. Provide clean in place connections and design for drainability so cleaning is quick when it comes. And be realistic in the design about the concentration at which your specific solution begins to deposit, rather than optimistic.
The vacuum system is half the plant
Evaporator offers are compared on the evaporator. A meaningful share of the capital, and a large share of the operating trouble, sits in the equipment that maintains the vacuum and condenses the vapour at the end of the train.
Vacuum is what allows the last effect to boil at a temperature low enough to be useful. Losing it does not stop the plant; it raises the boiling temperature, squeezes the temperature difference across every effect, and quietly reduces capacity while the operators look for a fault in the heating side.
Two arrangements are common. Steam ejectors have no moving parts, tolerate carryover and are extremely reliable, and they consume motive steam continuously, which is a real and permanent load on the utility system. Liquid ring vacuum pumps consume electricity instead and are frequently more economical to run, at the cost of a rotating machine and a seal liquid circuit that itself needs cooling.
The condenser sizing is the other item to interrogate. Surface condensers keep the vapour condensate separate from the cooling water, which matters if that condensate is recoverable as clean water or if the product carries anything that must not enter the cooling circuit. Direct contact condensers are cheaper and mix the two streams, which is fine until the condensate is contaminated and the cooling water becomes an effluent problem.
Non condensable gases are the detail that decides whether any of it works. Air leaking into a system running below atmospheric pressure accumulates in the condenser, blankets the heat transfer surface and destroys performance. Proper venting arrangements at every effect, and a plant designed to be leak tight in the first place, are the difference between a system that holds its vacuum and one that fights for it every shift.
Zero liquid discharge, and where the difficulty sits
Tightening discharge regulation has pushed a great deal of evaporation work into effluent treatment, and zero liquid discharge plants are now a substantial part of the field.
The engineering challenge is different from a product concentration duty. Effluent is variable in composition, frequently contains mixed and aggressive salts, scales readily, and is driven right to the point of crystallisation because that is the entire objective. Material selection has to be conservative because the liquor at the end of the train is far more aggressive than the feed. These plants combine evaporation with crystallisers and separation equipment, and they are best treated as an integrated system rather than a set of purchased items.
Where the payback actually sits
Compare two evaporator offers on capital cost alone and the smaller, simpler, lower economy plant will win every time. It will also spend the next fifteen years quietly consuming the difference in steam, several times over.
The honest comparison needs three numbers alongside the equipment price: steam consumption per tonne of water, electrical consumption per tonne of water, and expected run length between cleans. With those, the decision is straightforward arithmetic. Without them, you are comparing steel by weight and hoping the energy works itself out. It does not.
