Evaporators are used to turn the liquid form of a substance into its gaseous or vapour form.  

Historical Background

Evaporators are used in many processes besides refrigeration and air conditioning.  In particular, the chemical industries and food industries use them to evaporate water from solutions where more concentrated substances are required.  A multiple-effect evaporator, as defined in chemical engineering, is an apparatus for efficiently using the heat from steam to evaporate water.

The multiple-effect evaporator was invented by the American inventor and engineer, Norbert Rillieux.  The first industrially practical evaporator was built in 1845.  Originally designed for concentrating sugar in sugar cane juice, it has since become widely used in all industrial applications where large volumes of water must be evaporated, such as salt production and water desalination.  This type of evaporator uses the same underlying physical principles as described in the article below.


An evaporator is a device which turns the liquid form of a substance into its gaseous or vapour form.  This is a widely used process and the are many applications for it.  We restrict ourselves here to phase transitions between liquids and gases only, but similar physical laws apply for such transitions between solids (melting) and liquids or gases (sublimation)

Evaporation is a type of vaporisation that occurs on the surface of a liquid as it changes into the gas phase.  The surrounding gas must not be saturated with the evaporating substance.  When the molecules of the liquid collide, they transfer energy to each other.  When a molecule near the surface absorbs enough energy to overcome the vapour pressure, it will escape and enter the surrounding air as a gas.  When evaporation occurs, the energy removed from the vapourised liquid will reduce the temperature of the liquid, which in turn results in the phenomenon of evaporative cooling.

For those who are interested in more detail, we refer you to the Clausius-Clapeyron relation, named after Rudolph Clausius and Benoît Paul Émile Clapeyron.  The relation they found is a way of characterising a discontinuous phase transition between two phases of matter of a single constituent.  We will discuss this in a future article so please sign up to be notified when it is available.

What is vapour pressure?

Vapour pressure is the pressure exerted by a vapour when the vapour is in equilibrium with the liquid form of the same substance, i.e., when conditions are such that the substance can exist in both, or even in all three, phases.  Vapour pressure is a measure of the tendency of a material to change into the gaseous or vapour state,  The temperature at which the vapour pressure at the surface of a liquid becomes equal to the pressure exerted by the surroundings is called the boiling point of the liquid.

The vapour pressure of a liquid is the equilibrium pressure of a vapour above its liquid surface, i.e., the pressure of the vapour resulting from evaporation of a liquid above a sample of the liquid in a closed container.  The process of evaporation in a closed container will proceed until there are as many molecules returning to the liquid as the are escaping.  At this point the vapour is said to be saturated, and the pressure of that vapour is called the saturated vapour pressure.

Factors that affect resp. don’t affect vapour pressure are the following:

  • The surface of a liquid in contact with the gas does not affect the vapour pressure
  • The types of molecules that make up a liquid determine its vapour pressure, i.e., strong forces between molecules results in low vapour pressure, weak forces in high vapour pressure
  • At a higher temperature, more molecules have enough energy to escape from the liquid, resulting in higher vapour pressure

If the liquid is open to the air, then the vapour pressure is seen as a partial pressure along with the other constituents of the air.  The temperature at which the vapour pressure is equal to the atmospheric pressure is called the boiling point.

In order to evaporate, the liquid must collect a large heat of vapourisation, so evaporation is a potent cooling mechanism.  This is used in chillers, where evaporators are ultimately the component used to generate ‘cold’.  This is done by extracting heat energy from the medium to be cooled which is then used to evaporate the liquid refrigerant.

Sensible and latent heat

To understand the evaporation process, let’s first make sure we understand that heat, which is a form of energy, can exist on its own and can be moved from one place to another.  Heat always flows from warmer to colder places, a phenomenon known as the 2nd law of thermodynamics.

There are two kinds of heat – sensible heat and latent heat.

Sensible heat is literally the heat that can be felt.  It is the energy moving from one ‘system’ to another and that changes the temperature rather than changing its phase.  It is the heat that can be felt by standing next to a fire or outside in the sun, for example.  Sensible heat is the energy due to molecular motion, and changes in sensible heat are tangible because the temperature of the substance rises when the energy is supplemented to it.  The physical state of the substance however, will not be changed, i.e., from liquid to gas, because no boiling has occured.

Latent heat on the other hand, is the heat needed to change from one form of matter to another and which does not change the temperature of that matter.  Latent heat is the energy which is required for molecular separation and cannot be measured with a thermometer.  A latent heat change causes a change of physical state at a constant temperature.  The temperature does not change, but the state changes, i.e., from a liquid to a gas.

Types of evaporators and criteria for selection

As one of the most important components in a chiller, the purpose of the evaporator is to transfer the cooling energy from the cold gas of the refrigerant to water or another coolant.

There are three main types of evaporator for industrial chillers.  They are the coil type, shell and tube type and plate type.  We will examine these three types of evaporator from the aspect of heat transfer efficiency, requirements for water quality, ease of maintenance and space considerations.  Whilst we list coil type evaporators as one of the three types, they are not relevant for industrial chillers and so we mention them only for completeness.

Petra - Shell & Tube evaporator

Heat transfer efficiency

Heat transfer efficiency is very important for a heat exchanger.  Plate type heat exchangers, or evaporators, are more efficient than others due to their more precise structure.  Shell and tube type ranks second in this list.  Coil evaporators have the poorest levels of efficiency amongst the three types.  This is why they are not typically used in industrial chillers and why we do not discuss them further.

Heat efficiency needs to be based on the heat exchange area, higher efficiency means it can transfer more energy with less heat exchange area.  If the efficiency is lower, we need a larger heat exchange area.

Water quality

In general, plate evaporators and shell and tube evaporators require high water quality as on-site maintenance and remedial work is difficult.  With plate evaporators, the requirements are even greater as the water flow is very fine and can easily be restricted by foreign bodies.  When it comes to the corrosiveness of the water, all evaporator types must use anti-corrosion materials.

Space requirements

Why are the space requirements for the evaporator of interest to us?  Because for the same machine dimensions, the less space required for the evaporator means more space available for maintenance.  It may not be a crucial point, but nevertheless it is an aspect to consider.

The plate evaporator has the most precise structure compared to the other types already mentioned which means that it also offers the smallest size, whereas the shell and tube evaporator takes up the most space.  There are several types of shell and tube evaporator and large chillers can have several kilometres of tubing in their heat exchangers.

Chillers with direct evaporation of the refrigerant reserve the water to be cooled in the vessel and the liquid refrigerant is pushed through a large number of tubes.  During its journey through the tubes the refrigerant eventually becomes fully gaseous.  In this case, the refrigerant evaporates in the tubes.

Chillers with flooded evaporation reserve the refrigerant in the shell and the vessel is partly filled with layers of copper tubes that carry the water to be cooled down.  The level of the liquid refrigerant in the shell needs to be adjusted so that the top layer of tubes is just covered with liquid which is then evaporated if warmer water arrives in the tubes.  The flooded type of shell and tube evaporator is very efficient and the preferred solution for centrifugal chillers, but bears the disadvantage that a large amount of refrigerant is required.  Other types of evaporator need much less refrigerant.


About heat transfer in evaporators

The energetic processes in an evaporator can be assigned to two different areas.

The refrigerant absorbs the heat from the medium to be cooled and evaporates.  The temperature of the refrigerant remains constant despite heat absorption.  The absorbed energy is used to change the phase (latent energy).

The refrigerant, which has already evaporated completely, continues to absorb heat and heats up.  There is superheated refrigerant vapour in the outlet of the evaporator.  This so-called working superheat determines the degree of utilisation of the evaporator and can be adjusted via the expansion valve.

Petra - Shell & Tube evaporator

In the early days of industrial refrigeration, the majority of systems used flooded NH3.  Only later, when the so-called safety refrigerants (HCFC, HFC) were used, automatically controlled throttling devices became necessary.

Development and introduction of thermostatic expansion valves led to the question of whether or not an evaporator could be operated with an arbitrarily small superheat.

Flooded evaporators utilise the entire evaporator surface because of non-existent superheat.  In a dry expansion evaporator, valuable heat exchanger surface must be provided for superheating the refrigerant.

Note that only a little power is transferred in the superheat area of the heat exchanger because the heat transfer from gaseous refrigerant to the inner tube wall is much smaller than in the phase transition from liquid to gaseous refrigerant where latent heat transfer is the dominating effect.

It therefore makes sense to keep the superheat and the superheat surfaces as small as possible.  Too small a superheat can lead to unstable operation of the chiller and there is a risk of liquid ingression in the compressor.  Direct expansion evaporators are usually found to overheat between 6K and 10K, at which evaporators usually show stable operation.

The evaporation and superheating sections significantly influence the performance of an evaporator.  The diagram below shows the temperature curve of a dry evaporation process schematically.

To simplify matters, the evaporator in the diagram below was assumed to be an ideal countercurrent heat exchanger.  This simplification soes not represent a serious falsification of reality, since in the 2-phase range the temperature of the evaporating refrigerant is approximately constant, neglecting the pressure losses on the refrigerant side, and thus a distinction between counteflow, crossflow and co-current becomes obsolete.  The superheating section is sensibly placed at the air inlet of the evaporator in order to achieve the greatest possible temperature difference for a clear superheating.

Temperature differences

Several different temperature differences can be defined at an evaporator.

The ‘medium inlet temperature difference’ Δt1 is the difference between the medium inlet temperature tmedium_in and the evaporating temperature t0 at the outlet of the refrigerant.  Δt1 is used for the design of the evaporator, since medium inlet temperature and the evaporating temperature are typical design parameters and available from chiller design.

The mean temperature difference Δtmean is a theoretical value which represents the average of all temperatures differences over the entire heat exchanger surface. In the case of co-current or counter-current flow and constant heat transfers, it can be calculated with the known formula for the mean logarithmic temperature difference.

The mean temperature difference is of relevance for the heat exchanger manufacturer as it is influenced by the medium mass flow and the distribution of the heat transfer medium.

The medium cooling is the difference between the temperatures of the medium flow entering and leaving the evaporator.  It is strongly influenced by the medium velocity in the evaporator.  For the same inlet temperature difference Δt1, flooded evaporators have a larger mean temperature difference Δtm than direct expansion evaporators.

This is immediately apparent from the figure on the left, as the lack of superheat in flooded evaporators increases the area between the two temperatures curves.  However, the advantage of the larger average temperature difference is offset by the fact that the internal heat transfer coefficients are lower in flooded systems than in direct expansion systems.  Flooded evaporators also require a liquid separator downstream of the evaporator.  The liquid separator can be eliminated by ensuring that pure gaseous refrigerant reaches the compressor.  This is the task of the expansion valve.

We first consider the folllowing simple situation using a manually adjustable expansion valve, not a thermostatic one:

Evaporation temperature and media inlet temperature are kept constant.  Then the valve will almost be closed.  A high superheat now occurs in the evaporator which is only slightly filled with refrigerant and the cooling capacity is correspondingly low.

If the valve is opened further, more refrigerant enters the evaporator, the refrigerating capacity increases and the superheat decreases.  At a certain opening of the valve, a point is reached at which the superheat no longer assumes a constant value but begins to fluctuate.

This is a phenomenon of the evaporator only, not of the expansion valve.  If the valve is opened further and further, the oscillation disappears again, but the evaporator then runs ‘wet’ and liquid unevaporated refrigerant leaves the heat exchanger.  This can lead to compressor damage.

Slightly superheated suction gas always contains a high number of entrained droplets that are not in thermal equilibrium with the superheated gas and lead to noticeable efficiency degradation at the compressor.

Evaporator curves

Evaporator curves describe the behaviour of a heat exchanger and display its capacity as a function of superheat.  The red dot indicates the superheat temperature at which fluctuations start – see Δt1,a  on the figure top right.

Now we repeat the above procedure, keep the evaporation temperature constant, but vary the inlet temperature of the medium.  This leads to the other red dots in the figure below left.

We keep in mind: If we choose a lower medium temperature , the maximum possible superheat becomes smaller (right side of the curve) as well as the maximum possible evaporator capacity (left side of the curve).

Proceeding like this, with more superheat temperatures, we can identify a curve connecting all the red dots (below right).

This curve (above right) separates the plane into a stable and unstable region indicating the proper operation of the evaporator occurs best in the stable region.  Note that we would have found the same curve if we would have held the medium inlet temperature constant but varies the evaporation temperature

Now we understand that variation of Δt1 have a significant influence on the evaporator capacity. This is the so-called MSS theory. MSS stands for Minimum Stable Superheat and the red curve in the diagram above is named the evaporator stability curve.

In reality, and away from the bare theory, we know about other factors which impact on the evaporator performance like icing, contamination or differences in flow and flow rates.  All of these differences are completely independent from the type of expansion valve used in the system.

To even better understand the situation during cooling we move to a different presentation of the above diagrams i.e., we divide both quantities which make up the axes of the coordinate system by Δt1 and find new variables:


Horizontal axis = Δt1,aΔt1 dimensionless Vertical axis = evaporator capacity per Δt1 = Q0Δt1 in [kW/K]


The overheat level of an evaporator itself is defined as:

Δt0,superheatΔt1 = t0,superheatt0tmedium,int0


Let us consider one of the various evaporator curves and consider it as a ‘typical’ and ‘normalised’ curve.

The way to present this curve with the modified axis nomenclature is in line with EN328 which helps to determine the nominal capacity of evaporators.  In this standard, several conditions are defined which eventually cover a huge variety of evaporator curves.  The nominal capacity for each of these curves is always releated to an arbitrary (but well-proven) value of the overheat level i.e. 0.65.

Moving along such a curve to the left of 0.65 we risk crossing the MSS point which leads to increasing fluctuations, moving to the right of 0.65 we enter stable areas of operation which however, coincide with decreasing evaporator capacity.

We use 0.65 in our considerations because this value seems to be a good choice to work with.  This relation leads us quite naturally to the empirical values for superheat of direct expansion evaporators between 6K and 10K:

e.g. t0,superheat – t0 = 0.65 . (12°C – 3°C) = 5.9K for medium in = 12°C and evaporation tevap = 3°C
or   t0,superheat – t0 = 0.65 . (20°C – 6°C) = 9.1K for medium in = 20°C and evaporation tevap = 6°C


Direct evaporation and the interaction of evaporator and expansion valve have been the focus of attention for decades, especially for manufacturers of expansion valves.  Starting with automatic expansion valves, so-called constant pressure valves, the search was on for valves that filled the evaporator sufficiently with refrigerant, permitted stable control behaviour and reliably avoided liquid ingresion into the compressor.

Many experiments were carried out and these showed that the superheat temperature of an evaporator behaved stochastically i.e., after complete evaporation of the liquid refrigerant there is an unstable phase which only ends at some point of distance from the point of the MSS.

Due to the difficulty of locating the superheat as close as possible to this point, the development of electronic expansion valves began.  The aim was to achieve a control behaviour that keeps the characteristic curve of electronic expansion valves close to the current MSS line over the entire load range.  The controller itself must therefore recognise the MSS point in order to regulate the superheat in its vicinity.

For more on expansion valves, please see our page Expansion Valves

Other trends

In so called ‘falling film’ evaporators, evaporation takes place inside vertical tubes, but there are also applications where the process fluid evaporates on the outside of horizontal or vertical tubes.  In all cases, the process fluid flows downwards by gravity as a continuous film.  The fluid will create a film along the tube walls, progressing downwards (falling) – hence the name.

Falling film evaporators have a number of advantages over their flooded evaporator counterparts.  They require a lower refrigerant charge, as the entire shell (in the case of horizontal evaporators) or all the tubes (in the case of a vertical evaporator) need not be filled with liquid as a thin film is now used to cover the surfaces.  In industries such as heating and air conditioning this can save significant money due to the high cost of a refrigerant charge.

Falling film evaporators also show improved heat transfer characteristics over their flooded counterparts, particularly in cases with low heat flux.

A number of disadvantages exist, primarily the comparative lack of understanding of falling film evaporators compared to flooded evaporators, particularly for horizontal falling film evaporators.  Furthermore, the fluid distribution for horizontal falling film is a challenge as the performance is severely limited if an uneven distribution of film over the tubes is created.