Call Us
+86-18072289720
- All
- Product Name
- Product Keyword
- Product Model
- Product Summary
- Product Description
- Multi Field Search
Views: 0 Author: Site Editor Publish Time: 2026-08-18 Origin: Site
Search for how a compressor works and almost everything you find is about air conditioning: a house, a thermostat, a service van. The thermodynamics in those answers is not wrong. The frame around it is, and the frame is what decides whether a machine lasts fifteen years in a cold store or fails in its third season.
This page answers the question for industrial refrigeration — cold stores, blast freezers, process cooling, fish holds — where the compressor is doing exactly the same job but is chosen and killed by different numbers.
A semi-hermetic reciprocating machine: motor inside the refrigerant boundary, cylinders on top, service valves at the ends. Everything below happens between those two valves.
A compressor does not make cold. It maintains a pressure difference, and the pressure difference is what lets refrigerant boil below the temperature you need. Remove the compressor and the cold stops, but not because cold was being manufactured.
Two numbers decide whether the machine survives: pressure ratio and discharge temperature. Capacity gets the attention; those two get the warranty claims.
Four numbers describe a compressor properly. Displacement, capacity at a stated duty point, pressure ratio, discharge temperature. Seasonal efficiency ratings belong to comfort cooling, not to a freezer running at a fixed evaporating temperature.
"Two-stage" is ambiguous and people buy the wrong machine because of it. In air conditioning it usually means two capacity steps. In industrial refrigeration it means two stages of compression. They are not the same thing.
Refrigerant boils at a temperature set by its pressure. Drop the pressure far enough and it boils well below the temperature of the room you are trying to cool, which is why heat flows out of the room and into the refrigerant instead of the other way round. That is the whole trick. Everything else is plumbing, and the compressor is the pump that holds the pressure difference steady.
Four components, four states. The compressor's whole contribution is the left-hand leg: it takes vapour at suction pressure and delivers it at a pressure where the condenser can reject heat to whatever is available — sea water, ambient air, a cooling tower.
Read the loop once more, slowly, because two of the four steps are where plants actually go wrong.
Suction. The compressor draws vapour from the evaporator. It must be vapour: a compressor cannot pump liquid, and if liquid arrives the machine converts the energy of a moving piston into a hydraulic hammer instead of a pressure rise. The suction line is also what carries oil back to the crankcase, which is why its slope, its traps and its insulation are engineering decisions rather than installation details.
Compression. The volume closes and the pressure rises, and with it the temperature — not because heat was added, but because the same amount of energy now sits in a smaller space. This is the step that costs electricity and generates the discharge temperature that later decides how long the valves last.
Discharge and condensation. The hot high-pressure vapour enters the condenser, gives up heat to the cooling medium, and leaves as liquid. The temperature it can be cooled to is the single biggest lever on the whole plant's efficiency, and it is set by the condenser and the weather, not by the compressor.
Expansion. The liquid passes through a valve into the low-pressure side, flashes partly to vapour, and drops to the evaporating temperature. Cold is produced here, at the valve and in the evaporator — not in the compressor.
Ask what a compressor's capacity is and you will get a number in kW. Ask at what conditions and the answer sometimes gets vague. That vagueness is where projects go wrong, because a compressor has no capacity in the abstract: it has a capacity at an evaporating temperature, a condensing temperature and a refrigerant, and all three move it.
The two numbers that matter more than capacity are these.
Pressure ratio is discharge pressure divided by suction pressure, both absolute. It is the measure of how hard the machine is being asked to lift. As it climbs, volumetric efficiency falls — more of each stroke is lost to clearance gas re-expanding — and the discharge temperature climbs with it.
Discharge temperature is what actually destroys machines. Above a certain point the oil begins to break down, the valve plate erodes, and the clearances that the oil was sealing stop being sealed. A compressor running at a comfortable pressure ratio but a starved condenser can still cook itself.
Both numbers come out of the duty point, not out of a competitor's sales reply. If a supplier quotes capacity without stating the evaporating and condensing temperatures, the number is not comparable to anything.
Comfort cooling is rated on seasonal efficiency: a single number that folds a whole year of part-load operation into one figure, because an air conditioner spends most of its life at partial load and its load follows the weather.
An industrial refrigeration machine usually does not work that way. A blast freezer pulls down hard and then holds. A cold store runs near its design point for thousands of hours. A process chiller follows a product schedule. Seasonal ratings built around weather-driven part load describe almost none of that.
What to ask for | What it tells you |
|---|---|
Displacement, m³/h | How much volume the machine sweeps, and whether the frame is in the right class at all |
Capacity in kW at a stated evaporating and condensing temperature | The only capacity figure worth comparing |
Pressure ratio at that duty point | How hard the machine is working, and where the efficiency is going |
Discharge temperature at that duty point | Whether the oil and the valves will survive the duty |
Refrigerant and oil | Whether the machine can be serviced where it is installed |
Electrical data at the rated condition | Whether the supply and the protection are right |
Our own four-cylinder reciprocating range runs from 34.73 to 56.25 m³/h across eight models, 6 to 20 HP, on a 380–420 V three-phase supply. Those are the numbers a plant engineer sizes against; they are also the numbers a seasonal rating cannot express.
There are more architectures than four, but these are the ones that turn up in refrigeration plant rooms.
Reciprocating. Pistons, cylinders, suction and discharge valves. Positive displacement, so it keeps its capacity as pressure ratio rises, and it tolerates a wide range of refrigerants. Serviceable in the field: heads come off, valve plates are replaced, and the machine goes back to work. Ours are semi-hermetic, in two, four and six cylinder frames at Blait, and this is the architecture that covers most cold store and process duty up to a few hundred kW.
Screw. Two intermeshing rotors, continuous compression, far fewer moving parts and no suction or discharge valves to wear. It is the right answer at higher capacity and at high pressure ratio, which is why deep-freeze duty is screw territory. It is also less forgiving at very small capacity and costs more to make at small sizes.
Scroll. Two spirals, one orbiting. Excellent in packaged equipment and light commercial duty, where it is quiet and reliable. In industrial refrigeration it appears in small condensing units and in some cascade systems, not as the main machine on a freezer.
Centrifugal. An impeller throwing gas outward. Very large capacity, very large chillers, and a different set of problems — surge at low load being the one that breaks them.
Family | Where it wins | Where it loses |
|---|---|---|
Reciprocating | Wide duty range, high pressure ratio, field service, multi-refrigerant | More wearing parts; more maintenance attention per kW |
Screw | Large capacity, continuous duty, high pressure ratio, few wearing parts | Poor economics at small capacity; needs a real oil circuit |
Scroll | Small packaged systems, low noise, low maintenance | Capacity ceiling; not built for deep-freeze pressure ratios |
Centrifugal | Very large chilled water plant | Surge protection, narrow operating window, large minimum load |
Two of those four are what Blait builds: reciprocating in two, four and six cylinder frames, and screw in single-stage and single-body two-stage form. That is not a coincidence — it is the pair of architectures that covers industrial refrigeration from a small cold room to a deep-freeze plant.
This is the most expensive ambiguity in the industry, and it appears in specifications every week.
In air conditioning, "two-stage" usually describes capacity steps: a machine that runs at roughly seventy per cent or at full capacity, or a system with a two-stage thermostat. Nothing about the compression changes; the machine simply has two output levels instead of one.
In industrial refrigeration, "two-stage" describes two stages of compression: gas is compressed, cooled or de-superheated at an intermediate pressure, and compressed again. The reason is the pressure ratio — at a deep evaporating temperature a single stage would run at a ratio that no valve plate and no oil will tolerate for long.
The two meanings produce machines that look nothing alike, and a specification that says "two-stage" without saying which one cannot be quoted correctly. Our own two-stage screw is the second kind: two rotor sets in one body, motor between them, with a low-stage to high-stage displacement ratio of 2.18 to 1, and an intermediate pressure chamber doing the work that a two-machine arrangement does with an external vessel.
If you are writing a specification, write "two stages of compression, intercooled" or "two capacity steps". The seven extra words will save a redesign.
As the evaporating temperature falls, the pressure ratio rises, and three things happen together.
Volumetric efficiency falls, so a given displacement delivers less capacity and the frame gets bigger for the same duty. Discharge temperature rises, so the oil and the valves are being asked to survive more. And the mass flow falls, which means the oil has less gas to travel with — return from a long evaporator run becomes a design problem rather than a detail.
Somewhere below about minus thirty degrees the arithmetic stops favouring a single stage. The answer is either two stages of compression in one machine, or a two-stage reciprocating arrangement for the smaller end of the range. Which one is right is a function of capacity and of how the plant will be operated, not of preference.
Refrigerant choice sits on top of that. R404A and R507A remain the working HFCs in this band and are what most of our marine and cold store machines are built for. R22 survives in existing plant but is not a choice for a new one. CO2 transcritical is a different pressure regime entirely and a different conversation — a real one, but not a drop-in.
The compressor is the machine in the foreground. Everything behind it — separator, receiver, pipework, valves — is what turns a machine into a plant, and it is where most reliability is won or lost.
A compressor is one of roughly six things that have to be right. The other five:
The oil system — separator, cooler, level control, heater. On a screw this is half the installation; on a reciprocating machine it is the crankcase, the pump or splash system, and the heater.
The condenser — it sets the condensing temperature, and therefore the pressure ratio, and therefore everything the compressor has to do.
The evaporator and its feed — it sets the evaporating temperature and the superheat the compressor sees.
The controls — capacity control, protection, and the logic that decides when to unload.
The piping — line sizes, risers, traps, insulation. A correct compressor on incorrect piping is a warranty claim waiting for its first summer.
None of those five is improved by choosing a better compressor. All of them can destroy a good one. That is why the honest answer to "which compressor should I buy" is usually another question about the plant.
The failure list in a domestic air-conditioning article is a different list from the one in a plant room, because the machines and the duty are different. Four mechanisms account for most industrial failures.
Liquid slugging. Liquid reaches the suction and the compressor tries to compress it. Liquids do not compress, so the energy goes into the mechanicals: bent valves, damaged connecting rods, cracked heads. It usually starts with a feed valve set too wide, a defrost that returns liquid, or a suction line that lets oil and liquid accumulate.
Oil starvation. The oil leaves the compressor with the discharge gas, and it has to come back. When it does not — wrong line sizing, wrong slope, a trap that collects it, an evaporator that holds it — the bearings run dry long before anything looks wrong from outside. Oil pressure protection exists precisely because the failure is silent until it is total.
High discharge temperature. A starved condenser, an overcharged system or a duty point beyond the envelope pushes the discharge temperature up. The oil carbonises on the valve plates, the plates erode, capacity falls, and the machine runs hotter still. It is a slow cascade, and it starts with a condenser that nobody cleaned.
Burnout. In a semi-hermetic machine the motor sits in the refrigerant stream. When a winding fails it arcs, the arc breaks the refrigerant and the oil down into acid, and that acid travels the whole circuit. A replacement compressor dropped into that circuit fails within months unless the line set is flushed and suction filter driers are fitted. On an open-drive machine this particular failure mode does not exist, because there is no motor in the gas — a distinction worth knowing before accepting a diagnosis.
Under that head sits the valve plate — the wear item that decides a reciprocating machine's service interval. The blue component is an oil pressure switch, the protection that stands between a slow oil problem and a destroyed crankcase.
A compressor that has failed has usually been telling someone for months. These are the readings that produce a diagnosis rather than an opinion:
Suction and discharge pressures, converted to saturated temperatures for the refrigerant in use.
Suction and discharge temperatures at the same moment as the pressures.
Superheat at the evaporator outlet, and subcooling or approach at the condenser.
Current draw against the nameplate and against the commissioning record.
Oil level, oil colour, and whether the oil pressure switch has been tripping.
Run hours and the history of trip codes.
The comparison that matters is not against a generic table — it is against the commissioning record for that machine. A plant that has drifted two kelvin in condensing approach over three years is a plant with a fouled condenser, and no compressor change will fix it.
The rule of thumb that circulates in residential service — multiply the age by the repair cost and compare to a round number — assumes a domestic unit with a short life, a standard replacement and a homeowner paying retail. Almost none of that transfers to a plant room.
The industrial version of the decision has four questions, in this order.
Is the duty point still the right one? If the product mix changed, the store was extended or the freezing regime was tightened, the existing machine was selected for a plant that no longer exists. Replacing it like for like reproduces the problem.
Is the refrigerant still available and serviceable where the plant is? A machine on a refrigerant that is being phased down is a decision about the next decade, not about the next repair.
What is actually worn? Valve plates, gaskets, bearings and shaft seals are wear items and are meant to be replaced. A machine with a worn valve plate is not a failed machine; it is a machine due for a service.
What does downtime cost compared with the machine? In a cold store holding product, a week of lost capacity usually costs more than the compressor. That calculation, not a rule of thumb, is the one that decides.
The answer is sometimes neither repair nor replacement: it is a bigger frame, a second machine for pull-down, or a different architecture for the low stage.
Duty | Sensible starting point | The question to ask first |
|---|---|---|
Cold store, −10 to −25 °C | Semi-hermetic reciprocating, single stage | What is the worst-case condensing temperature in summer? |
Blast freezer, −35 to −45 °C | Two-stage: single-body screw, or two-stage reciprocating at the small end | How many pull-downs per day, and how long is each? |
Process cooling, above 0 °C | Single-stage screw or reciprocating | Is the load steady or does it swing hourly? |
Marine provision plant | Semi-hermetic reciprocating with class type approval | Which class, and does your flag accept it? |
Very large chilled water plant | Centrifugal | What is the minimum load, and how is surge handled? |
The pattern to notice is that the family follows the pressure ratio, and the pressure ratio follows the duty point. Everything else — brand, price, footprint — is a tie-breaker among machines that are already in the right class.
A refrigeration compressor holds a pressure difference so that a refrigerant can boil at a temperature below its surroundings. Everything that follows — displacement, pressure ratio, discharge temperature, oil, valves — is the price of holding that difference reliably for years.
The two things worth taking away are unglamorous. Get the duty point written down, both temperatures and the refrigerant, before anyone quotes a capacity. And when a specification says "two-stage", find out which two-stage it means.
If you want the next level of detail, the screw versus piston comparison covers how to choose between the two architectures we build, what is inside a semi-hermetic compressor walks through the reciprocating machine part by part, the two-stage screw article explains what the second stage actually does, and the step-by-step check is what to do when a machine starts misbehaving.
The ranges themselves: four-cylinder reciprocating, two-stage screw, two-stage reciprocating, and the CO2 range for transcritical work.
No. Cold is produced where the refrigerant evaporates, at the expansion valve and in the evaporator coil. The compressor maintains the low pressure that makes that evaporation happen at the temperature you need, and raises the vapour to a pressure where the condenser can get rid of the heat.
Because liquid is effectively incompressible. A piston or a rotor has nowhere to push it, so the force goes into the mechanical parts instead of into a pressure rise. The result is hydraulic shock: bent valves, damaged rods, cracked heads. Suction superheat and feed control exist to keep liquid out of the machine.
Two stages of compression: the gas is compressed, cooled or de-superheated at an intermediate pressure, then compressed again. It is used at high pressure ratios, which means low evaporating temperatures. In air conditioning the same phrase often means two capacity steps instead, which is a completely different machine.
At that evaporating temperature the pressure ratio rules out a single stage, so the choice is between a single-body two-stage screw and a two-stage reciprocating machine. The screw suits larger capacities and continuous duty; the reciprocating machine suits smaller capacities and frequent pull-downs.
Long enough that "years" is the wrong unit of measurement — a machine on scheduled service is measured in decades of operation, with valve plates, gaskets and seals replaced along the way. What shortens it is not age but duty: a condenser that is never cleaned, a duty point beyond the envelope, or an oil circuit that was never right.
Not for anything below about zero degrees. The pressure ratio, the refrigerant, the oil and the motor cooling are all specified for comfort cooling duty. Fitting one to a freezer produces a machine that runs outside its envelope from the first day, and the warranty will not survive the commissioning report.
