Agricultural Engineering · Post-Harvest, Food & Process Engineering · also Refrigeration and Air-conditioning, Cold Storage
Refrigeration cycles, equipment and cold-store design, from COP to the heat load of a room full of potatoes.
In the exams: GATE AG Section 7 (refrigeration COP/TR in 4 of 8 papers 2019–26; composite cold-store walls recur); ICAR AIEEA-PG Unit VI; ICAR AICE subject 38 Unit 6
Refrigeration is the removal of heat from a space or a product to hold it below the temperature of its surroundings. Heat does not flow from cold to hot on its own, so a refrigerating machine must be supplied with work (a compressor driven by a motor) or with high-temperature heat (an absorption machine driven by steam, hot water, waste heat or a flame). Air-conditioning is the related art of controlling the temperature, humidity, cleanliness and movement of air, and it rests on psychrometrics, the thermodynamics of moist air. Cold-chain engineering puts both to work on perishable produce: pre-cooling after harvest, cold storage, refrigerated transport and retail display, with the product held inside a safe temperature band at every link.
For an agricultural engineer the subject is practical. Fruits, vegetables, milk, meat, fish, eggs, flowers and seed are living or biologically active materials that lose weight, quality and value by the hour when warm.
Examiners worldwide test the subject mainly through calculation. Entrance and postgraduate papers (GATE Agricultural Engineering, ICAR AIEEA-PG, ICAR AICE-JRF/SRF, ASRB NET, and equivalents elsewhere) ask for short numerical answers: COP from enthalpies, tonnes of refrigeration, conduction through a composite cold-store wall, a product or respiration load, a freezing time by Plank's equation, a psychrometric mixing or bypass-factor calculation.
This manual follows the curriculum units in order. Property values quoted are typical textbook values for teaching; for design, take them from current handbooks, national standards and manufacturers' certified data, because they vary by product, cultivar, maturity, material and test method.
The second law of thermodynamics in its Clausius form says that heat cannot pass from a colder to a hotter body without some other effect. A refrigerator is the device that supplies that other effect. It absorbs heat Q_L from the cold space at temperature T_L, receives net work W, and rejects heat Q_H = Q_L + W to the surroundings at T_H.
The measure of merit is the coefficient of performance (COP), a ratio of the useful effect to the energy purchased. For a refrigerator COP_R = Q_L/W; for a heat pump COP_HP = Q_H/W = COP_R + 1. COP is not an efficiency and is normally greater than one. The upper limit is set by the reversed Carnot cycle operating between two reservoirs: COP_R,Carnot = T_L/(T_H − T_L), with temperatures in kelvin. Two consequences follow directly and are examined often. First, the COP falls as the temperature lift (T_H − T_L) grows, so a freezer at −25 °C rejecting heat to a 40 °C summer ambient is far costlier per kilowatt of cooling than a chiller at 5 °C. Second, raising the evaporating temperature by a degree helps more than lowering the condensing temperature by a degree, because T_L appears in the numerator as well as the denominator.
Practical machines use a working fluid, the refrigerant, which absorbs heat by boiling at low pressure and rejects it by condensing at high pressure.
The standard (ideal) vapour-compression cycle has four components and four processes, numbered from compressor suction.
Process 1–2, compression: dry saturated vapour leaves the evaporator and is compressed isentropically to condenser pressure. It leaves as superheated vapour. Compressor work per kilogram is w = h2 − h1.
Process 2–3, condensation: the vapour is desuperheated and then condensed at constant pressure to saturated liquid, rejecting q_H = h2 − h3 to air or water.
Process 3–4, expansion: the liquid passes through an expansion device (a capillary tube or valve) and its pressure drops to evaporator pressure. The process is a throttling one with no work and negligible heat exchange, so h4 = h3. Part of the liquid flashes to vapour, so state 4 is a low-quality wet mixture. Throttling is irreversible; it is the main reason the cycle falls short of Carnot, the other being the superheat horn at the compressor discharge.
Process 4–1, evaporation: the wet mixture boils at constant pressure and temperature, absorbing the refrigerating effect q_L = h1 − h4 from the space or product.
From these, COP = (h1 − h4)/(h2 − h1). The refrigerant mass flow needed for a capacity Q is m = Q/(h1 − h4), and the compressor power is P = m(h2 − h1). The condenser load is m(h2 − h3), which by the energy balance equals capacity plus compressor power. The ratio of heat rejected to heat absorbed, the heat rejection ratio, is therefore 1 + 1/COP; condensers are sized for it.
A real plant departs from the ideal cycle in predictable ways. Vapour leaves the evaporator a few kelvin superheated, deliberately, so that no liquid reaches the compressor. Liquid leaves the condenser a few kelvin subcooled, which lowers h4 and increases the refrigerating effect. Compression is not isentropic: the isentropic efficiency (h2s − h1)/(h2 − h1) is below one, so actual work and discharge temperature are higher. Pressure drops in lines and valves add to the pressure ratio. A liquid-to-suction heat exchanger subcools the liquid and superheats the suction gas; it helps COP with fluids such as R-134a but gives little or no gain with ammonia.
The pressure ratio across one compressor rises steeply as evaporating temperature falls, lowering volumetric efficiency and raising discharge temperature. When it becomes too large, commonly above about eight to ten for reciprocating machines, compression is split into two stages, with the intermediate pressure near the geometric mean of suction and discharge pressures. A flash intercooler between stages removes flash gas and desuperheats the first-stage discharge. A cascade system uses two circuits with different refrigerants, the condenser of the low-temperature circuit being the evaporator of the high-temperature one; carbon dioxide below ammonia is a common modern pairing for freezing.
Refrigerating capacity is a rate of heat removal. The SI unit is the kilowatt, but the tonne of refrigeration (TR) persists in trade, examinations and Indian cold-storage practice. It was the rate of heat absorption that would melt one short ton (2000 lb) of ice at 0 °C in 24 hours: with latent heat 144 Btu/lb, 288,000 Btu per day or 12,000 Btu/h. In SI, 1 TR = 3.517 kW, about 211 kJ/min; some older texts use 210 kJ/min. TR is not compressor power, and it is not the capacity to make one tonne of ice a day from warm water, which also needs the water's sensible heat and losses to be removed.
The energy efficiency ratio (EER), capacity in Btu/h per watt of input, equals 3.412 × COP. Industry also quotes specific power in kW per TR; COP = 3.517/(kW per TR), so 0.7 kW/TR is a COP of about 5. The cycle COP uses work on the refrigerant; an overall COP divides by electrical input and so includes motor and mechanical losses.
For a given refrigerant and saturation temperatures, the refrigerating effect fixes the mass flow, and mass flow times suction specific volume fixes the volume a positive-displacement compressor must move. The volumetric refrigerating effect (h1 − h4)/v1 in kJ/m³ therefore decides compressor size; carbon dioxide's is very high because it works at high pressure. Capacity of a given compressor falls as evaporating temperature falls (less dense suction gas, lower volumetric efficiency) and as condensing temperature rises, so every rating must name its operating temperatures.
The pressure–enthalpy (p–h, or Mollier) chart is the working chart of refrigeration because three of the four processes of the cycle plot as straight lines on it. Pressure is plotted vertically on a logarithmic scale and specific enthalpy horizontally. The saturation dome divides the chart: to its left is subcooled liquid, inside it wet mixture, to its right superheated vapour. The peak of the dome is the critical point. Lines of constant temperature are horizontal inside the dome (because saturation temperature is fixed by pressure), nearly vertical in the liquid region and curve downward in the vapour region.
On this chart the ideal cycle plots as follows. Evaporation 4–1 is a horizontal line at low pressure ending on the saturated-vapour line. Compression 1–2 follows a constant-entropy line up into the superheated region. Condensation 2–3 is a horizontal line at high pressure ending on the saturated-liquid line. Throttling 3–4 is a vertical line straight down, since enthalpy is constant. The horizontal lengths read directly as energies: 4–1 is the refrigerating effect, 1–2 the compressor work, 2–3 the heat rejected.
Lowering the evaporating pressure moves line 4–1 down. The throttling line 3–4 lengthens, so point 4 moves right into higher quality and the refrigerating effect shrinks slightly; the isentropic line from a lower starting point is longer, so work rises; and the suction specific volume increases. The COP falls and the capacity of a given compressor falls sharply.
Raising the condensing pressure moves line 2–3 up. Point 3 moves right along the saturated-liquid line, so h4 rises and the refrigerating effect falls; the compression line lengthens, so work rises. COP falls. Dirty condensers, air trapped in the condenser (non-condensable gases) and hot weather all raise head pressure in this way.
Subcooling moves point 3 left into the liquid region, lowering h4 and increasing the refrigerating effect at no extra compressor work. Useful superheat, gained inside the refrigerated space, adds to the refrigerating effect; useless superheat, gained in suction lines outside the space, raises specific volume and work without useful cooling.
For a pure refrigerant a horizontal line inside the dome is also an isotherm. For a zeotropic blend (the R-400 series) temperature changes during evaporation and condensation at constant pressure; this change is the temperature glide, and the chart shows sloping isotherms inside the dome.
A refrigerant cycle has these state enthalpies read from a chart: saturated vapour at evaporator outlet h1 = 395 kJ/kg, compressor discharge after isentropic compression h2 = 425 kJ/kg, saturated liquid at condenser outlet h3 = 250 kJ/kg. Throttling gives h4 = 250 kJ/kg. The refrigerating effect is 395 − 250 = 145 kJ/kg and the work 425 − 395 = 30 kJ/kg, so the COP is 145/30 = 4.83. For a plant of 10 TR, capacity is 35.17 kW; the mass flow is 35.17/145 = 0.243 kg/s; compressor power is 0.243 × 30 = 7.28 kW; heat rejected in the condenser is 0.243 × (425 − 250) = 42.4 kW, which equals 35.17 + 7.28 within rounding.
An absorption system replaces the mechanical compressor with a thermal one. The condenser, expansion device and evaporator are as in a vapour-compression plant. Low-pressure vapour from the evaporator is absorbed into a liquid absorbent in the absorber, a process that releases heat and must be cooled. The strong solution (rich in refrigerant) is pumped to high pressure; pumping a liquid needs very little work compared with compressing a vapour. In the generator heat is supplied and the refrigerant is driven out of solution as high-pressure vapour, which goes to the condenser. The weak solution returns to the absorber through a pressure-reducing valve, usually after giving up heat to the strong solution in a solution heat exchanger.
In the ammonia–water system ammonia is the refrigerant and water the absorbent. Because water is itself volatile, the vapour leaving the generator carries water vapour, which would freeze or accumulate in the evaporator. An analyser and a rectifier (a dephlegmator) cool the vapour and return the condensed water to the generator.
The three-fluid (diffusion) absorption refrigerator uses ammonia, water and hydrogen, and is silent with no moving parts. The hydrogen equalises total pressure throughout the circuit so that no pump is needed; the refrigerant evaporates by diffusing into the hydrogen at low partial pressure.
In the lithium bromide–water system water is the refrigerant and the salt solution the absorbent. The salt is non-volatile, so no rectifier is needed. Because water freezes at 0 °C, these machines are used only above freezing, as chillers for air-conditioning and process cooling, typically producing chilled water in the range of about 5–10 °C. The whole low-pressure side runs under deep vacuum, so leak-tightness is critical. The main operating hazard is crystallisation: if the solution becomes too concentrated or too cold, salt crystallises and blocks the circuit, commonly in the solution heat exchanger. Controls limit the solution concentration, and corrosion inhibitors are added because lithium bromide solution is corrosive to steel in the presence of air.
Single-effect lithium bromide machines driven by low-pressure steam or hot water have COPs commonly around 0.6–0.8. Double-effect machines, which use the vapour from a high-temperature generator to heat a second generator, reach roughly 1.0–1.3 but need higher driving temperatures. These are typical ranges, not guarantees.
An absorption machine exchanges heat with three reservoirs: it receives heat Q_G at generator temperature T_G, absorbs Q_E at evaporator temperature T_E, and rejects heat at an intermediate temperature T_0 (the absorber and condenser). Its COP is defined as Q_E/Q_G, with pump work neglected. The ideal (reversible) limit is the product of a Carnot engine efficiency operating between T_G and T_0 and a Carnot refrigerator COP between T_E and T_0:
COP_max = [(T_G − T_0)/T_G] × [T_E/(T_0 − T_E)].
With a generator at 100 °C, surroundings at 30 °C and evaporator at −10 °C, COP_max = (70/373.15) × (263.15/40) = 1.23. Real single-effect machines achieve roughly half such values or less. The comparison with a vapour-compression COP of 3–5 is misleading unless the cost and source of the energy are compared: absorption uses low-grade heat, which may be free.
Absorption is attractive where waste heat, biogas, solar thermal energy or cheap steam is available, where electricity is scarce or expensive, and where silence or absence of vibration matters. Its drawbacks are size, weight, first cost, the larger cooling-water load (heat rejected equals Q_E + Q_G) and slower response.
Refrigerants are designated by ASHRAE Standard 34 and the matching ISO 817. For the halocarbons of the methane and ethane series, the number R-XYZ is read as X = number of carbon atoms minus one (omitted when zero), Y = hydrogen atoms plus one, Z = fluorine atoms; remaining bonds are chlorine (or bromine, shown by a B suffix). Thus dichlorodifluoromethane CCl2F2 is R-12, chlorodifluoromethane CHClF2 is R-22, and 1,1,1,2-tetrafluoroethane CH2FCF3 is R-134a, the lower-case letter distinguishing isomers. Hydrocarbons follow the same rule where possible: propane C3H8 is R-290. Isobutane is R-600a, from the 600 series for miscellaneous organic compounds. Inorganic compounds are the 700 series, the last two digits being the molecular mass: ammonia R-717, water R-718, carbon dioxide R-744. Zeotropic blends are numbered in the 400 series (R-404A, R-407C, R-410A) and azeotropic blends in the 500 series (R-502, R-507A).
By chemistry, the families are chlorofluorocarbons (CFCs, fully halogenated with chlorine, such as R-11 and R-12), hydrochlorofluorocarbons (HCFCs, which contain hydrogen and chlorine, such as R-22), hydrofluorocarbons (HFCs, no chlorine, such as R-134a, R-32 and blends), hydrofluoroolefins (HFOs, unsaturated with a carbon–carbon double bond, such as R-1234yf and R-1234ze(E)), and the natural refrigerants ammonia, carbon dioxide, hydrocarbons, water and air.
Ozone depletion potential (ODP) compares a substance's ability to destroy stratospheric ozone with that of R-11, which is assigned 1.0. Chlorine and bromine atoms released in the stratosphere catalyse ozone destruction; fluorine does not. CFCs have high ODP; HCFCs have much smaller ODP because the hydrogen makes them break down mostly in the lower atmosphere; HFCs, HFOs and natural refrigerants have zero ODP.
Global warming potential (GWP) compares the radiative forcing of a mass of gas over a time horizon, normally 100 years, with that of the same mass of carbon dioxide (GWP = 1). Many HFCs have GWPs of hundreds to several thousand; ammonia has zero, hydrocarbons a few, and HFOs very low values. Published GWP values change with each IPCC assessment report, and regulations specify which report's values apply, so always quote the source. For example R-134a is 1430 on the 100-year values of the Fourth Assessment Report used in several regulations, and R-404A about 3900 on the same basis. The Montreal Protocol (1987) phased out CFCs and is phasing out HCFCs; the Kigali Amendment (2016) brought HFCs under a phase-down; regional rules such as the European F-gas regulation go further.
Total equivalent warming impact (TEWI) adds the direct effect of refrigerant leaked over the plant's life (charge × leak rate × life × GWP, plus end-of-life loss) to the indirect effect of the carbon dioxide emitted in generating the electricity used. For many plants the indirect term dominates, so a low-GWP refrigerant with a poor COP is not automatically better for the climate.
ASHRAE 34 and ISO 817 classify refrigerants by toxicity (class A lower, B higher) and flammability (1 no flame propagation, 2L lower flammability with low burning velocity, 2 flammable, 3 higher flammability). Most HFCs are A1; R-32 and the HFOs R-1234yf and R-1234ze(E) are A2L; hydrocarbons such as propane and isobutane are A3; ammonia is B2L. Carbon dioxide is A1 but works at very high pressure, and its critical temperature of about 31 °C means that in warm climates the high side operates above the critical point (a transcritical cycle) with a gas cooler rather than a condenser.
Ammonia remains the workhorse of large cold stores, dairies and freezing plants. It has excellent thermodynamic properties, a high latent heat, good heat transfer, zero ODP and GWP, it is cheap, and its pungent smell reveals leaks at concentrations far below dangerous levels. Against it: it is toxic, it is flammable within a limited range, and it attacks copper and its alloys, so ammonia systems are built in steel and use open or specially designed semi-hermetic compressors rather than copper-wound hermetic ones.
Selection weighs thermodynamic performance at the design temperatures, pressure levels (neither deep vacuum nor excessive pressure), safety class and the charge limits that codes set for occupied spaces, compatibility with oil and materials, environmental regulation now and over the plant's life, cost and the availability of trained technicians.
The compressor raises the refrigerant from evaporator to condenser pressure and circulates it. Positive-displacement compressors trap a volume of gas and reduce it: reciprocating, rolling-piston and rotary-vane, scroll and screw types. Dynamic compressors, chiefly centrifugal, raise pressure by giving the gas velocity and converting it to pressure in a diffuser.
Reciprocating compressors cover small hermetic units in refrigerators up to large multi-cylinder industrial machines. Rolling-piston rotaries are common in room air-conditioners. Scroll compressors, with one fixed and one orbiting spiral, are quiet and efficient in the range of small commercial and air-conditioning plant. Screw compressors, twin-rotor or single-screw, dominate medium and large industrial refrigeration including ammonia cold stores; they are oil-injected to seal, cool and lubricate, tolerate some liquid, and have a built-in volume ratio that should be matched to the operating pressure ratio. Centrifugal compressors handle very large volume flows and are used in large water chillers; they are sensitive to surge, an unstable flow reversal at low flow and high head.
By enclosure, a hermetic compressor has motor and compressor sealed in a welded shell, cooled by suction gas; it cannot be opened for repair. A semi-hermetic compressor has a bolted shell that can be opened. An open compressor has an external shaft driven by a separate motor through a shaft seal; it allows any motor or engine drive and is required for ammonia where copper windings would be attacked.
At the end of the discharge stroke a small clearance volume V_c of high-pressure gas remains in the cylinder. On the suction stroke this gas re-expands before the suction valve can open, so less fresh gas is drawn in than the swept volume V_s. With clearance ratio C = V_c/V_s and polytropic re-expansion index n, the clearance volumetric efficiency is
η_v = 1 + C − C (p_d/p_s)^(1/n).
For C = 0.05, pressure ratio 5 and n = 1.2, η_v = 1 + 0.05 − 0.05 × 5^(0.833) ≈ 0.86. Volumetric efficiency falls as pressure ratio rises and would reach zero at a limiting ratio where the re-expanding gas fills the whole cylinder; this is one reason for staging at low evaporating temperatures. The actual volumetric efficiency is lower still because of suction-gas heating, valve pressure drops and leakage.
The swept volume rate of a reciprocating compressor is (π/4) D² L N k, for bore D, stroke L, speed N and k cylinders (for single-acting machines). Actual suction volume flow = η_v × swept volume rate, and mass flow = actual suction volume flow/v1.
Refrigeration loads vary with season, product entry and time of day, and the compressor must follow. Methods include on–off cycling (simple but hard on motors and poor for temperature stability), cylinder unloading in multi-cylinder reciprocating machines, slide-valve control in screw compressors, inlet guide vanes in centrifugal machines, multiple compressors in parallel sequenced by suction pressure, and variable-speed drives, which usually give the best part-load efficiency.
Protective devices include high-pressure and low-pressure cut-outs, an oil-pressure failure switch, motor overload and thermal protection, discharge-temperature limits and crankcase heaters to prevent refrigerant migrating into the oil during off cycles.
The condenser rejects the heat absorbed in the evaporator plus the compressor work. Three regions exist in it: desuperheating of the discharge gas, condensation at constant temperature, and a little subcooling of the liquid. Most of the heat is latent.
Air-cooled condensers use finned coils and fans. They need no water, suit small and medium plant and remote sites, but condense at a higher temperature than water-cooled units in hot weather, which lowers COP. Water-cooled condensers, most often shell-and-tube with water in the tubes and refrigerant in the shell, give lower condensing temperatures and compact size, but need a supply of clean water, usually recirculated through a cooling tower, with water treatment against scale, corrosion and biological growth including Legionella. Evaporative condensers spray water over a coil carrying the refrigerant while a fan draws air through; the evaporation of water gives condensing temperatures close to the ambient wet-bulb temperature and uses much less water than a once-through water-cooled condenser. They are standard in large ammonia plants.
Two cooling-tower terms are examined often. Range is the drop in water temperature across the tower (inlet minus outlet). Approach is the difference between the cold-water outlet temperature and the ambient wet-bulb temperature; a tower can never cool water below the wet-bulb temperature.
In the evaporator refrigerant boils and absorbs heat. By the way refrigerant is fed, evaporators are dry-expansion (direct-expansion, DX), in which the expansion valve meters just enough liquid to evaporate completely and leave a little superheat, or flooded, in which the evaporator is kept full of liquid from a surge drum and vapour is separated and returned to the compressor. Flooded evaporators wet the whole surface and transfer heat better. In liquid-overfeed (pumped recirculation) systems a pump sends several times the evaporated mass of liquid through the coils; this is common in large ammonia cold stores because it gives high coil efficiency and easy control of many rooms.
By construction there are bare-tube coils, finned-tube coils (forced-air unit coolers in cold rooms), plate evaporators (plate freezers, ice-bank tanks), and shell-and-tube chillers for liquids.
The evaporator temperature difference (TD), the difference between room air temperature and the refrigerant evaporating temperature, decides both coil size and room humidity. A large TD gives a small coil but removes much moisture from the air as the air is cooled far below its dew point, drying the room and the produce. Fresh fruits and vegetables that must not wilt are held with small TDs, often a few kelvin; designers choose the TD for the product.
Frost on the coil insulates it and blocks airflow. Defrost methods are off-cycle (for rooms above about 2 °C, air in the room melts the frost while the coil is off), electric heaters, hot gas from the compressor discharge, and water spray. The defrost heat that enters the room is part of the room load.
Condenser and evaporator are heat exchangers and are sized from Q = U A ΔT_m, where ΔT_m is the log-mean temperature difference. When the refrigerant side is at constant temperature (condensation or evaporation of a pure fluid), the flow arrangement does not matter and ΔT_m = (ΔT_1 − ΔT_2)/ln(ΔT_1/ΔT_2), with ΔT_1 and ΔT_2 the temperature differences at the two ends. For water entering a condenser at 30 °C and leaving at 35 °C with refrigerant condensing at 40 °C, the end differences are 10 and 5 K and ΔT_m = 5/ln 2 = 7.2 K.
The expansion device drops the liquid from condenser to evaporator pressure and meters flow to match the load; overfeeding floods the evaporator and risks liquid at the compressor, underfeeding starves it. The capillary tube is a long, small-bore copper tube. Friction and acceleration of the flashing refrigerant produce the pressure drop. It has no moving parts, is cheap and lets pressures equalise during the off cycle so that the compressor restarts unloaded, which allows a low-starting-torque motor. It does not respond to load, so it suits systems with steady loads, such as domestic refrigerators, freezers and room air-conditioners. Such systems are critically charged: the refrigerant charge must be correct within narrow limits and there is no liquid receiver.
The automatic, or constant-pressure, expansion valve holds evaporator pressure constant. Its behaviour is the reverse of what a load change needs: when load rises and evaporator pressure tends to rise, it closes, starving the evaporator. It is used only for small, steady loads.
The thermostatic expansion valve (TXV, TEV) holds a nearly constant superheat at the evaporator outlet. A sensing bulb clamped to the suction line contains a charge whose pressure acts on the top of a diaphragm; evaporator pressure and an adjustable spring act below. When load rises, superheat rises, bulb pressure rises and the valve opens. Where the evaporator has a large pressure drop, an external equaliser line from the evaporator outlet to the underside of the diaphragm is needed; otherwise the valve would sense a higher pressure than exists at the outlet and would hold excessive superheat.
Float valves control liquid level instead of superheat. A low-side float valve holds the level in a flooded evaporator or surge drum; a high-side float valve passes to the evaporator whatever liquid the condenser produces, so the evaporator holds most of the charge. Room temperature is usually held by a thermostat switching a liquid-line solenoid valve; with pump-down, the compressor empties the evaporator to a low-pressure cut-out before stopping, so liquid cannot migrate to it.
Atmospheric air is treated as a mixture of dry air and water vapour, each behaving as an ideal gas. The humidity ratio (specific humidity) ω is the mass of water vapour per kilogram of dry air: ω = 0.622 p_v/(p − p_v), where p_v is the partial pressure of vapour and p the total pressure; 0.622 is the ratio of the molar masses of water and dry air. Relative humidity φ = p_v/p_s, the ratio of the vapour pressure to the saturation pressure at the same dry-bulb temperature. The dew-point temperature is the saturation temperature at the actual vapour pressure; a surface colder than it collects condensation. The wet-bulb temperature is read by a thermometer with a wetted wick in moving air; for air–water mixtures it is close to the adiabatic saturation temperature.
Enthalpy of moist air per kilogram of dry air is h = 1.005 t + ω (2501 + 1.88 t) kJ/kg, with t in °C: the first term is the sensible heat of dry air, the second the latent heat plus sensible heat of the vapour. All psychrometric quantities are per kilogram of dry air because the mass of dry air is conserved through air-conditioning processes, while the mass of vapour is not.
On the psychrometric chart, dry-bulb temperature is the horizontal axis and humidity ratio the vertical axis. The processes are:
Sensible heating or cooling: a horizontal line; ω is constant. Q = m_a (h2 − h1), or approximately m_a × 1.02 × Δt using the humid specific heat about 1.005 + 1.88ω kJ/kg K.
Cooling and dehumidification: air passes over a coil whose surface is below the dew point. The air leaves cooler and drier. The process line points toward the apparatus dew point (ADP), the effective coil surface temperature on the saturation curve. The coil bypass factor (BF) is the fraction of air treated as if it passed untouched: BF = (t_out − ADP)/(t_in − ADP); contact factor = 1 − BF.
Heating and humidification: winter conditioning, heating followed by adding steam or water spray.
Evaporative (adiabatic) cooling: air passes over wet pads or through a spray; sensible heat evaporates water, so temperature falls and humidity rises along a line of nearly constant wet-bulb temperature. Saturation efficiency = (t_in − t_out)/(t_in − t_wb). It works well in hot, dry climates and is widely used for low-cost produce storage structures and greenhouses; it cannot cool below the wet-bulb temperature.
Adiabatic mixing of two streams: the mixed state lies on the straight line joining the two states, divided in inverse proportion to the dry-air masses. Mixed humidity ratio and enthalpy are mass-weighted averages; dry-bulb temperature is very nearly so.
Chemical dehumidification: air passes over a desiccant; humidity falls and temperature rises, close to constant enthalpy.
The room sensible heat factor (SHF) is sensible heat gain divided by total (sensible plus latent) heat gain. The line through the room condition with the slope given by the SHF is the condition line; supply air must lie on it to absorb both loads in the right proportion. A low SHF (large latent load, for example from wet produce or many occupants) requires supply air cooled to a low dew point, often with reheat. For a supply volume flow, the sensible load is Q_s = m_a c_p Δt and the latent load Q_L = m_a Δω h_fg, with h_fg about 2450–2500 kJ/kg. Cooling-coil load = m_a (h_in − h_out) minus the small enthalpy of condensate removed, often neglected.
In cold-store design psychrometrics enters three ways: the enthalpy difference between outside and inside air sets the infiltration load; the evaporator TD sets the room humidity; and the dew point of outside air sets where condensation and ice will form on and inside the envelope.
The refrigeration load of a cold store is the sum of heat gains over a design day. The usual components are: transmission through walls, roof and floor; product load (sensible heat removed in cooling the product, latent heat in freezing it, sensible heat below freezing); respiration heat from living produce; infiltration of warm air through doors; and internal loads from people, lights, fan motors, forklifts and defrost. Packaging and pallets are cooled too and are included with the product. The components are added, a safety factor (commonly about 10%, set by the designer) is applied, and the 24-hour total is divided by the planned daily running time of the plant, often 16–20 hours, to size the equipment.
Heat conducts through each surface at Q = U A ΔT, where U is the overall heat-transfer coefficient of the construction, A the area and ΔT the difference between outside design temperature and room temperature. For a composite wall of layers in series, 1/U = 1/h_o + Σ(x_i/k_i) + 1/h_i, where h_o and h_i are outside and inside surface coefficients, x the thickness and k the thermal conductivity of each layer. In a well-insulated store the insulation dominates: 150 mm of polyurethane with k = 0.025 W/m K has a resistance of 6 m² K/W, beside which the surface resistances (roughly 0.03–0.13 m² K/W each) are small.
Walls exposed to sun gain extra heat; designers add a sun-effect temperature allowance to ΔT for sunlit walls and roofs, or shade them. The floor ΔT is taken from ground temperature, which is lower than peak air temperature. Under freezer floors the ground itself would eventually freeze; the water in it expands into ice lenses that heave and crack the floor, so freezer floors have underfloor heating (warm-air ducts, glycol pipes or electric cables) or a ventilated void beneath the insulation.
Product sensible heat above freezing is Q = m c_p (t_in − t_f) where t_f is either the storage temperature (for chilled goods) or the initial freezing point. Latent heat of freezing is Q = m L; for foods, L is roughly the water mass fraction times the latent heat of water, 334 kJ/kg. Sensible heat below freezing uses the frozen specific heat. Where measured values are not given, Siebel's relations estimate specific heat from water mass fraction x_w: above freezing c_p ≈ 3.35 x_w + 0.84 kJ/kg K; below freezing c_p ≈ 1.26 x_w + 0.84 kJ/kg K. These are approximations; foods do not freeze at a single temperature, and property data from handbooks should be used for design.
The product load must be converted to a rate using the time allowed for cooling. A store that receives 20 tonnes of potatoes a day at 30 °C to be cooled to 4 °C within 24 hours has a very different load from a holding room already full of cold potatoes. Loading a whole store in a few days at field heat is the single commonest cause of undersized plant in developing-country cold stores.
Harvested fruits, vegetables, flowers and tubers are alive. They respire, oxidising sugars to carbon dioxide and water and releasing heat. The heat of respiration depends strongly on the commodity and on temperature, rising roughly two- to threefold for each 10 K increase over the usual storage range, and it varies with cultivar, maturity and injury. Handbooks tabulate it, commonly in watts or milliwatts per kilogram or per tonne at several temperatures. Load is Q = m × (respiration rate at the storage temperature).
Respiration heat is small per kilogram compared with the field heat that must be removed on entry, but it continues every hour for the whole storage life and scales with the mass stored, so in a large holding store it is a significant steady load. Frozen and non-living products such as meat, frozen fish and dairy produce have no respiration load.
Every door opening exchanges cold, dense room air for warm, moist outside air. The warm air must be cooled, and its moisture condensed or frozen on the evaporator. Two estimating methods are used. The air-change method takes a number of room volumes of outside air per 24 hours, tabulated against room volume and use (small rooms have more air changes per unit volume than large ones, and heavy traffic raises the figure). Load = air changes per day × room volume × (h_out − h_in)/v_out, with enthalpies in kJ/kg dry air and v_out the specific volume of outside air, or approximately using outside air density. The door-flow method estimates the air exchanged by density-driven flow through an open doorway from door size, temperature difference and opening time; it is preferred for large doors and busy rooms.
Infiltration is usually the largest load in a busy distribution store. Strip curtains, air curtains, rapid-roll doors, vestibules and anterooms, and loading docks sealed to the truck body cut it substantially.
People give off heat at a rate that rises as the room gets colder, because they work harder and wear more clothing that is itself losing heat; published tables for cold stores give values rising from roughly 200 W per person at 10 °C to around 400 W at −25 °C. Load = number of people × heat per person × hours present.
Lighting load is the installed wattage times hours of use; all electrical energy for lights ends up as heat in the room. Motors inside the room, chiefly evaporator fans, add their full electrical input as heat if the motor is in the room; if only the driven machine is in the room the shaft power is the heat. Battery forklifts and conveyors add their power. Electric or hot-gas defrost adds a fraction of the defrost energy.
The time to freeze a product is needed to size freezers and to schedule loading. Plank's equation, derived by treating the product as initially at its freezing temperature, only latent heat being removed, constant properties of the frozen layer and a sharp freezing front, is
t_F = (ρ L/(T_f − T_a)) × (P a/h + R a²/k),
where ρ is density of the frozen product, L the latent heat of the product (kJ/kg, converted to J/kg), T_f the initial freezing temperature, T_a the cooling-medium temperature, h the surface heat-transfer coefficient, k the thermal conductivity of the frozen product, and a the characteristic dimension: thickness of an infinite slab, diameter of an infinite cylinder or of a sphere. The shape constants are P = 1/2, R = 1/8 for a slab; P = 1/4, R = 1/16 for an infinite cylinder; P = 1/6, R = 1/24 for a sphere.
The first term in the bracket is the surface (convective) resistance and the second the internal (conductive) resistance. The equation shows that freezing time rises with the square of thickness when internal resistance dominates but only linearly when the surface coefficient is low, so faster air or immersion helps thin products most. Because it ignores precooling and subcooling and assumes a sharp freezing point, Plank's equation underestimates real freezing times; modified forms (such as Cleland and Earle's) and numerical methods correct this. For examinations, use the equation as given with consistent SI units and convert seconds to hours.
The insulation fixes the transmission load for the life of the building. Good cold-store insulation has low thermal conductivity, low water-vapour permeability, low water absorption, adequate compressive strength for floors, dimensional stability at low temperature, acceptable fire behaviour, and it is free of odour and vermin attraction. The common materials are rigid polyurethane (PUR) and polyisocyanurate (PIR) foams, usually as factory-made sandwich panels with steel faces, with thermal conductivity typically about 0.022–0.028 W/m K; expanded polystyrene (EPS), about 0.033–0.040; extruded polystyrene (XPS), about 0.029–0.035, preferred under floors for its strength and low water uptake; mineral wool and glass wool, used where fire rules demand non-combustible insulation; and historically cork and rice-husk or sawdust in low-cost stores. Values vary with density, blowing agent, age and temperature; use certified data.
The economic thickness of insulation is the thickness that minimises the sum of annual capital charges on insulation and the annual cost of the energy to remove the heat that leaks through it. It is greater for freezers than for chill rooms and rises with electricity price. For a given allowable heat flux q, thickness x = k ΔT/q, neglecting surface resistances.
Water vapour moves from high to low vapour pressure, which in a cold store is from the warm outside toward the cold inside, for most of the year and all day. As the vapour passes into the insulation it meets progressively colder layers. Where the temperature falls below the local dew point, the vapour condenses; below 0 °C it freezes. Wet insulation conducts heat far better than dry, and ice builds up year after year, eventually destroying the insulation and sometimes the structure.
The remedy is a continuous vapour barrier (vapour retarder) of very low permeance on the warm side of the insulation, with all joints, penetrations and fixings sealed. Steel-faced sandwich panels provide the barrier in their outer skin, and the joints are the weak point. The cold side should be relatively more permeable so that any moisture that does enter can escape into the room and be removed by the evaporator. A barrier on the cold side, or on both sides with leaks on the warm side, traps moisture.
A second check is surface condensation on the outside skin: the outside surface temperature must stay above the outside air dew point. It can be found from the resistance ratio, t_surface = t_out − (R_outside film/R_total)(t_out − t_in). Thin insulation or thermal bridges such as steel columns, door frames and fixings can bring surfaces below dew point.
Freezer rooms and sealed cold stores need pressure-relief ports: when warm air enters during a door opening and then cools, it contracts and can pull a strong vacuum on the envelope; defrost heating can do the reverse. Floors are built with a load-bearing slab over insulation with underfloor heating. Many national guidelines, such as India's NCCD engineering guidelines for cold stores, set minimum system standards for insulation, refrigeration and safety for subsidised projects; consult the current national document.
An ice bank is a form of thermal energy storage. A tank of water contains evaporator coils or plates. During periods of low demand, often overnight or between milkings, a relatively small refrigeration unit runs continuously and builds a layer of ice on the evaporator surface. When cooling is needed, chilled water at close to 0 °C is pumped from the tank through a plate or tubular heat exchanger, or sprayed over the outside of a milk vat, and returns warmer to melt the ice. The latent heat of fusion of ice, about 334 kJ/kg, stores a large amount of cooling in a small mass: each tonne of ice stores about 334 MJ, roughly 93 kWh of cooling.
The advantage is that the compressor is sized for the average daily load rather than the peak; the bank also rides through short power cuts, and water at 0 °C cannot freeze the milk. A dairy that must cool a large batch of warm milk within a short time after milking would otherwise need a compressor sized for that short peak. With an ice bank the compressor runs for many hours to store cooling and the stored ice supplies the peak.
Milk leaves the udder at around 35–37 °C and bacteria multiply rapidly at that temperature. Regulations in most countries require raw milk to be cooled to a low temperature, commonly 4–6 °C, within a set time after milking; the exact limits vary by country and by whether the milk is collected daily. Bulk milk coolers are either direct-expansion, with the evaporator bonded to the tank wall, or ice-bank types. Direct-expansion coolers are more efficient in energy because they evaporate at a higher temperature and have no intermediate water loop, but need a larger compressor and a reliable power supply at milking. Ice-bank coolers suit sites with weak or intermittent power, such as village milk collection centres in many developing countries, and large herds with short milking peaks.
The ice-bank capacity calculation is an energy balance. Sensible heat to cool milk: Q = m c_p Δt, with c_p for whole milk about 3.9 kJ/kg K. Mass of ice required = Q/L_ice (adding a margin for losses). The compressor must build that mass of ice in the time available, and it works at a low evaporating temperature, so its COP is lower than a DX cooler's. Ice-thickness controls, either electronic sensors or a probe that detects the change in conductivity between water and ice, stop the compressor when the design thickness is reached; excessive ice bridges between coils and slows heat transfer.
Eutectic plates, sealed containers of a salt solution with a freezing point below 0 °C, store cooling in refrigerated vehicles and insulated containers: they are frozen overnight from a mains-powered unit and absorb heat during the delivery round. Phase-change materials with melting points chosen for the product (for example near 5 °C for vaccines or milk, or well below zero for frozen food) are used in passive shipping boxes.
A cold chain is the sequence of temperature-controlled steps from harvest or slaughter to the consumer: pre-cooling, pack-house handling, cold storage, refrigerated transport, distribution centre, retail display and home storage. The chain is only as good as its weakest link: produce warmed on an unrefrigerated loading dock for two hours may lose more shelf-life than it gains from a week in a perfect store.
Pre-cooling removes field heat quickly after harvest, before storage or transport. Methods are room cooling (slow, simply placing produce in a cold room), forced-air cooling (fans pull cold air through vented packages stacked with a tarpaulin to force the air through, several times faster than room cooling), hydrocooling (immersion in or showering with chilled water, suited to produce that tolerates wetting such as many root vegetables and some fruits), vacuum cooling (lowering pressure so water evaporates from the produce, very fast for leafy vegetables with a high surface-to-mass ratio, with about 1% weight loss per 5–6 K of cooling), package icing and top icing for tolerant commodities such as broccoli. Cooling rate is described by the half-cooling time, the time to reduce the temperature difference between product and cooling medium by half. Because cooling is roughly exponential, seven-eighths cooling time, the practical end point for many operations, is three half-cooling times.
Optimum temperature and relative humidity differ by commodity. Most temperate fruits and vegetables keep best near 0 °C and 90–95% relative humidity. Many tropical and subtropical products, including bananas, mangoes, many cucurbits and tomatoes, suffer chilling injury (pitting, discoloration, failure to ripen, decay) when held below a critical temperature that is well above freezing, commonly in the range of about 7–13 °C for the most sensitive. Potatoes for seed are held cold, but potatoes for processing into chips and fries are held warmer, because at low temperatures starch converts to reducing sugars that darken on frying (cold-induced sweetening). Onions and garlic need low humidity. Ethylene producers, such as ripening apples, bananas and tomatoes, should not share a room or container with ethylene-sensitive commodities such as leafy greens, carrots and cut flowers. Standard compatibility tables group commodities by temperature, humidity and ethylene; they are guides, not guarantees.
Controlled-atmosphere (CA) storage lowers oxygen and raises carbon dioxide to slow respiration, extending apple storage to many months. It needs gas-tight rooms and careful control, since too little oxygen causes anaerobic respiration and off-flavours.
Road vehicles use insulated bodies with mechanical refrigeration units driven by a diesel engine, by the vehicle engine, or electrically from a generator, battery or standby mains; eutectic plates for multi-drop delivery; or cryogenic systems that release liquid nitrogen or carbon dioxide into the body or into a heat exchanger. Refrigerated containers (reefers) for sea and rail carry their own electrically driven unit, powered from the ship or a generator set; air is delivered under the load through a grooved (T-bar) floor and returns across the top, so correct stowage with no blocking of the floor and no loading above the red line is essential. Reefers are designed to maintain temperature, not to pull down warm cargo, so products must be pre-cooled to carriage temperature before loading.
International road and rail carriage of perishable foodstuffs in many countries is governed by the United Nations ATP Agreement, which defines classes of insulated, refrigerated and mechanically refrigerated equipment and requires testing and certification. Its insulation classes are normal insulated equipment, with overall heat-transfer coefficient K not exceeding 0.70 W/m² K, and heavily insulated equipment, K not exceeding 0.40 W/m² K. The heat leak of a body is K × mean surface area × temperature difference, and the refrigeration unit must exceed it by a margin that the agreement specifies for certification.
Temperature must be recorded, not assumed. Data loggers in loads and in rooms, remote telemetry on reefer units and time-temperature indicators on packages provide evidence for food-safety management systems based on HACCP principles and for claims against carriers.
In developing countries post-harvest losses of perishables are high, and cold storage is often concentrated on a few commodities such as potatoes, while pre-cooling and refrigerated transport links are weak. Small solar-powered cold rooms, evaporative coolers for short-term holding, ice-bank milk chilling centres and pack-house pre-coolers are practical interventions; their success depends as much on operation, maintenance and business model as on engineering.
Learn to draw the cycle on a p–h chart without looking and to write every energy term in kJ/kg before touching capacity. Practise converting TR to kW both ways, and building a cold-store load sheet line by line: transmission for each surface, product sensible and latent heat, respiration, infiltration, people, lights, fans, defrost, safety factor, run hours. Most exam errors are unit slips, so write units at every step. In the laboratory, measure COP on a trainer by energy balance on the evaporator water or air and compare it with the chart value; the difference teaches you where real cycles lose.
A young engineer sizes plant for pack-houses, potato stores and milk chilling centres, commissions systems and troubleshoots. Typical decisions: how many tonnes per day of field-hot produce the store can take, whether to add a forced-air pre-cooler, what evaporator TD gives the humidity the product needs, how thick the panels should be, and whether an ice bank or direct-expansion cooler suits a village's power supply. On site you read suction and discharge pressures and superheat to diagnose undercharge, overcharge, air in the condenser, blocked driers and iced coils, and you log temperatures to prove the chain was kept.
A senior engineer owns whole cold-chain projects: master-planning stores, pre-coolers, docks and reefer logistics; writing specifications and checking contractors' load calculations; choosing refrigerant strategy under the HFC phase-down, weighing ammonia's efficiency and safety management against low-charge or CO2 systems; negotiating energy tariffs and thermal storage; and setting maintenance, safety and emergency procedures for ammonia plant under national codes. Such an engineer also investigates failures, such as iced insulation, floor heave, chilling injury claims and product losses, and defends conclusions to insurers and regulators.
A researcher with long field experience works on the questions standard tables do not answer: respiration and property data for local cultivars, validated freezing and cooling models for real packages and stacks, airflow uniformity in stores and reefers by computational fluid dynamics checked against measurements, low-GWP and natural-refrigerant systems for hot climates, solar and thermal-storage cold rooms for off-grid farms, and the economics of loss reduction along whole value chains. The work links engineering to food science and to policy, and must report uncertainty honestly because design decisions rest on it.
This manual does not reproduce refrigerant property tables or p–h charts, commodity-by-commodity storage and respiration data, or door-flow infiltration formulae in full; take those from the ASHRAE Refrigeration handbook, USDA Handbook 66 and national guidelines. It gives only an outline of transcritical CO2 cycle analysis, ammonia plant safety codes (which differ by country), controlled-atmosphere design, refrigerated container operation and the detailed rules of the ATP certification tests. Regulatory phase-down dates for HCFCs and HFCs are not listed because they differ by country group and are amended; check the current Montreal Protocol schedules and national rules. Typical property ranges quoted (insulation conductivity, absorption COPs, chilling-injury thresholds, heat per person) are teaching values that vary with product, material and source.
Original practice questions written to the syllabus and pattern of the exams named in each question's examTags. They are not past papers and do not reproduce any board's questions. Every numerical answer was computed by script from the values given in the question; property values in questions are illustrative and vary by product and material.
These notes are free to read here, with no account.
What the Laisha app adds:
Free for ten days, then a subscription. We would rather write that here than have you find it later.