Energy efficiency in industrial refrigeration – where do losses occur and how can they be reduced?

In industrial refrigeration, it is easy to focus on the compressor because it is usually the largest single energy consumer. That view is too narrow, however. The cost of refrigeration is created by the whole system: compressors, condensers, evaporators, pumps, fans, the defrost system, auxiliary installations and controls, as well as process conditions and the weather.

That is why the energy efficiency of a cold store or refrigeration system should be assessed under actual operating conditions. It is necessary to know not only how much energy the compressor uses, but also the evaporating and condensing temperatures at which it operates, how much useful cooling it actually delivers to the process and how much energy is consumed by auxiliary equipment. This system-level approach is also recommended by the U.S. Department of Energy in its industrial refrigeration guidance.

In brief: where should you start when assessing a refrigeration system?

• First, establish the actual process requirements: product temperature, room temperature, chilled-water or glycol temperature, and acceptable deviations.
• Measure the energy use of the entire system, not just the compressors.
• Check evaporating and condensing temperatures. An excessive difference between them increases compression work.
• Assess how the system behaves at part load and how multiple compressors, pumps and fans are staged.
• Before replacing major equipment, check heat exchangers, flows, setpoints, sensors, defrosting, leak tightness and controls.
• Treat free cooling and heat recovery as solutions that work together with the refrigeration system. Their economic viability must be calculated for the balance of the whole system.

Industrial refrigeration efficiency must be calculated for the whole system

A compressor datasheet may specify COP, the ratio of cooling capacity to input power. This is important information, but it does not yet describe the efficiency of the refrigeration system at the facility. The official UK Energy Technology List also defines COP as cooling capacity divided by power input and assesses equipment at strictly defined evaporating and condensing temperature points. These test conditions show why a catalogue value cannot be considered independently of operating temperatures.

In a real facility, the compressor operates together with other loads. The electrical energy balance should therefore include, among other things:

• compressor motors and their capacity-control systems,
• condenser and evaporator fans,
• chilled-water, glycol and condenser-water pumps,
• cooling towers or evaporative condensers, where present,
• defrost energy, including electric heaters,
• drain-pan heaters, door heaters or other auxiliary loads if they belong to the system being analysed,
• controls and additional equipment where their consumption is material.

Only then can an indicator for the entire system be calculated. In simple terms:

System COP = useful cooling capacity / total electrical power of the refrigeration system.

If cooling is transferred by water or a glycol solution, useful cooling capacity can be determined from the flow rate, temperature difference and fluid properties. In direct-expansion systems, measurement is more difficult and may require data from controllers, meters, the process balance or equipment performance models.

From an auditor’s perspective, the first mistake is often defining the balance boundary too narrowly. If only the compressors are measured, the system may appear efficient even though pumps, fans or poorly controlled defrosting are consuming excessive energy.

Where do energy losses most often occur in industrial refrigeration?

1. Evaporating temperature is too low

The compressor must raise the refrigerant pressure from the evaporating side to the condensing side. The lower the evaporating temperature at the same condensing temperature, the greater the temperature and pressure difference the system has to overcome.

The evaporating temperature should therefore be kept as high as practicable while still meeting process requirements. This does not mean simply increasing the cold-room temperature setpoint. It is necessary to determine the temperature actually required by the product or process, the temperature difference across the evaporator and whether the heat exchanger has adequate air or secondary-fluid flow.

An unnecessarily low evaporating temperature may result, among other things, from:

• an excessively low chilled-water, glycol or room temperature setpoint,
• a dirty or iced evaporator,
• restricted airflow through the air cooler,
• incorrect secondary-fluid flow,
• an excessive safety margin in the setpoints,
• combining loads with very different temperature requirements in one common circuit.

This is why separating circuits operating at different temperature levels is an important optimisation measure. A load requiring a very low temperature should not unnecessarily impose those conditions on the entire system.

2. Condensing temperature is too high

On the high-pressure side, the problem is the opposite: the higher the condensing temperature, the more work the compressor has to perform. The cause may be not only a hot day, but also the technical condition of the equipment and the way the system is controlled.

It is worth checking:

• the cleanliness of fins and heat-transfer surfaces,
• airflow through the condenser,
• fan operation and staging sequence,
• condenser-water temperature and flow where a water-cooled system is used,
• the condition of the cooling tower or evaporative condenser,
• scale or other fouling on the water side,
• recirculation of hot air around the condenser,
• the minimum condensing-pressure setpoint and whether it can vary with outdoor temperature.

Many systems can use so-called floating head pressure, i.e. reduce condensing pressure when outdoor conditions and equipment requirements allow it. However, it should not be adjusted without analysing the minimum operating requirements of expansion valves, the oil system and the manufacturer’s recommendations. The DOE identifies reducing condensing pressure and using a floating setpoint as fundamental measures for improving industrial refrigeration efficiency.

3. Inefficient compressor operation at part load

A facility rarely needs its maximum refrigeration capacity throughout the year. Load varies with production, outdoor temperature, cold-room door openings, product temperature and the operating schedule. Assessing the system only at full load can therefore lead to incorrect conclusions.

It is necessary to check how the compressors control capacity and what their actual electrical power is at 25, 50, 75 and 100% load, where such data are available. In practice, manufacturer performance maps, controller data or measurements should be used. It must not be assumed that part-load efficiency changes linearly.

Warning signs can include frequent starts and stops, long periods of unloaded operation, large fluctuations in suction pressure or situations where several machines operate simultaneously at low load. In its refrigeration and process-cooling guidance, the DOE identifies both appropriate compressor staging and speed control as potential optimisation measures, although the correct solution depends on the characteristics of the specific system. Its process-cooling guidance emphasises that compressors, pumps, fans and controls should operate as one system.

4. Poor sequencing of multiple units operating in parallel

In plant rooms with several compressors, it is not enough to ask, “Which compressor is the most efficient?” The relevant question is which combination of machines gives the lowest power demand for the required cooling load and the current operating pressures.

The same issue applies to pumps and fans. Two pumps running in parallel are not always better than one. Conversely, switching off an auxiliary device purely to reduce its own consumption may worsen heat transfer and increase compressor energy use. Optimisation must take the balance of the entire system into account.

Good control should therefore track load, pressures, temperatures and the efficiency of different equipment combinations. It is also worth checking whether the controls keep too many units running “just in case”.

5. Dirty, frosted or poorly loaded heat exchangers

The evaporator and condenser determine the temperatures between which the compressor operates. Deteriorating heat transfer increases the temperature difference required to transfer the same amount of energy.

On the evaporator, the problem may be dust, grease, ice or frost. On the condenser, it may be dust, dirty fins, water-side scale or insufficient flow. For frost, DOE/ORNL technical literature identifies two adverse effects: increased airflow resistance and increased thermal resistance of the ice layer. The result is restricted heat transfer and lower heat-exchanger efficiency.

Before deciding to replace a compressor or refrigeration unit, it is therefore worth measuring temperatures upstream and downstream of the heat exchanger, the temperature difference between the secondary fluid and the refrigerant, flow rate and pressure drop. These data often show that the source of the problem lies outside the compressor.

6. Fans and pumps operating at constant output

Auxiliary energy use in a refrigeration system can go unnoticed because it is distributed across many motors. If pumps continuously maintain maximum flow while valves throttle the excess, or if fans run at full speed regardless of load, it is worth assessing whether the control method can be changed.

The analysis should include at least:

• motor power and speed,
• fluid flow rate,
• available pressure or differential pressure,
• control-valve positions,
• variable-speed drive setpoint signals,
• supply and return temperatures.

A variable-speed drive does not automatically guarantee savings. It is necessary to check whether reducing flow will impair heat transfer, disrupt fluid distribution or increase compressor work.

7. Incorrect defrosting

Defrosting has two opposing energy costs. If it is carried out too often or for too long, unnecessary heat is introduced into the room and the heat exchanger. If it is carried out too infrequently, growing frost restricts airflow and heat transfer.

It is therefore worth recording not only when defrosting starts, but also:

• cycle duration,
• the method used to terminate the cycle,
• heat-exchanger temperature,
• room temperature before and after defrosting,
• power draw of the heaters, if used,
• time required to return to normal conditions,
• the number of door openings and moisture sources if they materially affect frost formation.

The best strategy depends on the type of system, product, humidity and operating practice. Changing the schedule without measurements may simply shift the problem from energy use to storage quality or air-cooler performance.

8. Setpoints and controls that do not follow the process

Many losses are not caused by a fault. The system may be operating exactly according to its programmed logic, but that logic may have been configured for different production conditions or may never have been optimised for energy performance.

Typical problems include:

• fixed setpoints regardless of load and weather,
• deadbands that are too narrow or too wide,
• controllers that “fight” each other over temperature or pressure,
• manual overrides left in place after servicing,
• sensors with measurement error or offset,
• incorrect staging of compressors, pumps and fans,
• failure to reduce condensing pressure in cooler conditions,
• maintaining an unnecessarily low fluid temperature around the clock,

Trends from the control system are particularly useful here. A single instantaneous value often looks correct. Only a record covering several hours or days reveals oscillations, frequent switching or equipment operating when the process does not require full capacity.

9. Outdoor conditions and variable process load

Refrigeration energy use should not be compared solely month to month. Higher outdoor temperatures increase condensing conditions in air-cooled systems, while wet-bulb temperature is also important for evaporative condensers. In cold stores, the temperature and humidity of air entering through doors matter, while in production processes the quantity, temperature and timing of incoming product are important.

If production increases by 20%, an increase in refrigeration energy use alone does not necessarily indicate lower efficiency. Conversely, constant production does not mean constant refrigeration load if product temperature or outdoor conditions have changed. Energy data should therefore be normalised against the variables that actually affect cooling demand.

What data should be collected when assessing a refrigeration system?

A good audit does not start with a catalogue for a new chiller or refrigeration unit, but with data. The measurement scope depends on the system, but the table below shows a set of parameters that will usually make it possible to identify the cause of losses.

Data resolution should be sufficient to show equipment operating cycles. In a rapidly changing system, hourly averages may hide short starts, unloading periods and oscillations. Conversely, a single day of measurements may not capture other production shifts or weather conditions. The logging range and frequency should therefore be selected to match the dynamics of the system rather than a single universal rule.

How should efficiency be measured? COP is only the starting point

COP of the entire system

If useful cooling capacity can be measured, it should be compared with the power demand of all equipment used to produce it. This gives an indicator for the actual system rather than for a single machine.

Assume that a system produces chilled water at a flow rate of 50 m³/h, and the temperature difference between return and supply is 4 K. Using, for this example, a water density of 1000 kg/m³ and a specific heat capacity of 4.18 kJ/(kg·K), the cooling capacity is approximately:

50 / 3600 × 1000 × 4.18 × 4 ≈ 232 kW of cooling.

If the entire system is drawing 80 kW of electrical power at that moment, then:

System COP ≈ 232 / 80 = 2.9.

If only the 65 kW drawn by the compressors were included, the calculated COP would be about 3.6. The difference is not a mathematical error; it comes from excluding 15 kW of auxiliary equipment. In a real glycol system, the actual density and specific heat for the relevant concentration and temperature must be used.

Indicators that help identify the cause

COP alone does not explain why a system is performing worse. It is useful to compare it with other indicators, including:

• evaporating and condensing temperatures and the difference between them,
• temperature differences across heat exchangers,
• electrical power per unit of useful cooling produced,
• kWh per tonne of product or another production unit, if it properly represents the load,
• compressor operating hours and number of starts,
• share of operating time at individual load levels,
• pump and fan power relative to the entire system,
• energy use for defrosting,
• efficiency under comparable outdoor and production conditions.

Two COP values should not be compared without checking the measurement conditions. The same machine can have very different efficiency at different evaporating and condensing temperatures.

What symptoms may indicate inefficient operation?

Not every symptom below automatically means a fault, but each is a good reason to examine the data:

• condensing temperature remains high even on cooler days,
• evaporating temperature is clearly lower than required by the process temperature,
• compressors frequently switch between loaded and unloaded operation or cycle on and off frequently,
• several compressors run in parallel at low load,
• fans or pumps continuously run at full speed despite variable demand,
• valves remain heavily throttled for most of the time while pumps maintain high pressure,
• the evaporator frosts up quickly or defrosting starts more often than the actual level of icing would justify,
• the temperature difference across a heat exchanger increases at the same output,
• energy use increases despite similar production and comparable outdoor conditions,
• the system does not adjust setpoints to weather or load,
• operators regularly leave controls in manual mode because the automation cannot maintain stable conditions,
• room or process temperature is consistently below the value actually required by the process.

The relationship between symptoms is more valuable than any single symptom. For example, rising condensing temperature, fans at full speed and an increasing temperature difference across a dirty condenser form a much stronger clue than an increase in compressor power alone.

What should be checked before replacing a compressor or refrigeration unit?

Replacing the main machine may be justified, but it should be the end of the diagnostic process, not the beginning. Before making an investment decision, it is worth working through operational and low-cost measures in sequence.

1. Verify temperature requirements. Determine the temperature the product or process actually needs and whether all loads must operate at the same temperature level.
2. Check heat-exchanger cleanliness. Clean condensers and evaporators in accordance with technical requirements and assess whether temperatures and pressures change after cleaning.
3. Restore correct flows. Check filters, dampers, valves, pumps, fans and heat-exchanger passages.
4. Verify sensors and controls. A temperature or pressure measurement error may force the entire system to operate harder than necessary.
5. Check defrosting. Compare the schedule with actual icing, cycle duration and the effect on room temperature.
6. Analyse compressor sequencing. Check performance maps, part-load levels, number of starts and how additional units are brought online.
7. Optimise condensing pressure and fluid setpoints. Introduce variable setpoints only where the equipment and process allow it.
8. Check pumps and fans. Assess the possibility of speed control, eliminating throttling or bypasses, and matching flow to demand.
9. In cold stores, check insulation, doors and air infiltration. Damaged seals and doors left open for long periods increase both the heat load and moisture ingress.
10. Measure after changes. Only a before-and-after comparison under similar load conditions will show how much of the problem was caused by operation and how much remains attributable to the equipment itself.

If, after these measures, the compressor still operates outside its efficient range, has mismatched capacity, an unsuitable operating profile or a technical condition that justifies modernisation, replacement options can be compared on a sound basis. The new machine should then be selected using the annual load profile, not only the maximum design point.

In practice, I would first look for the reasons why the existing system has to operate with such a large temperature lift and with so many units running. Only then would I assess whether compressor efficiency is genuinely the main limitation.

System tightness and refrigerant

Refrigerant charge and system tightness are important from both technical and operational perspectives. If the system uses fluorinated greenhouse gases, the obligations arising from Regulation (EU) 2024/573 must also be checked. The Regulation provides, among other things, for leak checks for specified types and quantities of refrigerants and for record-keeping requirements. The exact obligation depends on the refrigerant, its quantity and the type of equipment, so an energy-efficiency assessment does not replace an assessment of compliance with servicing requirements.

Free cooling: when can it reduce the load on mechanical refrigeration?

Free cooling uses low outdoor temperatures to cover some or all of the cooling demand without operating a refrigeration compressor. In process systems, for example, this may mean cooling water or glycol through a dry cooler or economiser when outdoor temperature is sufficiently low.

Assessing the potential requires comparing:

• the required cooling-fluid temperature,
• the minimum temperature difference required across the heat exchanger,
• the annual outdoor-temperature profile,
• the cooling-demand profile,
• pump and fan power in free-cooling mode,
• the cost and hydraulic constraints of the additional heat exchanger.

Free cooling usually has greater potential in circuits with higher cooling temperatures than in systems requiring deep freezing. It should therefore not be added to a project as an automatic saving; the number of hours in which conditions genuinely allow compressor load to be reduced must be calculated. The DOE also identifies the use of water-side economisers as one of the measures for process-cooling systems.

Heat recovery from refrigeration: a benefit only when the balance of the whole system improves

A refrigeration system removes heat from the process and rejects it on the condenser side together with the energy supplied to the compressor. This creates a natural source of waste heat that can be used, for example, to heat water or another process.

However, condensing temperature should not be artificially increased merely to obtain a hotter medium without checking the effect on compressor energy use. A good heat-recovery option should reduce total site energy consumption rather than shift the cost from the boiler room to the refrigeration plant room.

A broader approach to evaluating the heat source, heat sink, time profile and temperatures is described separately in our article Heat recovery in industry.

What should an energy efficiency audit of a refrigeration system include?

A robust assessment is not based on a single power measurement. It should combine documentation, historical data and measurements under different operating conditions.

1. Define the system boundary and function

First, map which equipment produces cooling, which equipment transports it, which loads receive it and what the process requirements are. Groups with different evaporating temperatures should be treated separately where they actually operate at different levels.

2. Energy and cooling balance

Available data on cooling capacity, flows, temperatures and production load are matched to the electrical energy-demand profile. Where submeters are unavailable, temporary measurements on compressors, pumps and fans are often required.

3. Record operating conditions

At the same time, pressures, temperatures, speeds, on/off states, control signals and outdoor conditions should be recorded. This makes it possible to link an increase in energy use to a specific cause.

4. Analyse part-load operation and control

The system is assessed at low, medium and high demand. Equipment sequencing, pressure stability, floating setpoints, variable-speed drive operation and frequent switching are particularly important.

5. List measures from operational to capital-intensive

Start by costing setpoint corrections, servicing, cleaning, calibration, schedule changes and changes to control logic. Only then should larger upgrades be compared: heat exchangers, new compressors, separation of circuits, controls, free cooling or heat recovery.

6. Verify the effect

Every improvement should have a baseline and a method for post-implementation measurement. For larger projects, analysis of the refrigeration system should be incorporated into the broader scope of energy audits and measurements, so that it can be assessed together with the site’s other significant energy users and processes.

What does this type of analysis mean for a facility?

The key conclusion is simple: industrial refrigeration should not be assessed solely on compressor efficiency. High energy use may result from an unnecessarily high compression ratio, dirty heat exchangers, poor sequencing, excessive flows, defrosting or setpoints that do not reflect current production conditions.

A well-executed diagnostic assessment should distinguish three groups of measures:

1. operational and maintenance changes that can be implemented without replacing major equipment,
2. changes to controls and automation that improve operation of the entire system under variable load,
3. investments whose expected effect is supported by measurement and analysis of the annual operating profile.

If a facility is planning a refrigeration upgrade, it is worth starting with measurements and reconstructing the actual operating profile. Such an audit separates operational issues from equipment limitations and only then provides a sound basis for comparing investment options.

FAQ – energy efficiency of refrigeration equipment and systems

How can the energy efficiency of refrigeration equipment be assessed?

Reading the compressor COP from a datasheet is not enough. Useful cooling capacity must be compared with the energy use of the entire system, while evaporating and condensing temperatures, part-load operation, pump and fan operation, defrosting and outdoor conditions must also be checked. Only this type of measurement shows system efficiency in the actual process.

Does replacing a refrigeration compressor always deliver the greatest savings?

No. If the compressor operates with suction pressure that is too low, condensing pressure that is too high or poor sequencing, a new machine may still operate under unfavourable conditions. The causes that increase the load on the whole system should be addressed first, and only then should machines be compared.

Which matters more: evaporating temperature or condensing temperature?

Both. Compressor work depends on the difference in conditions between the low- and high-pressure sides. Opportunities are therefore sought simultaneously to raise evaporating temperature and lower condensing temperature, provided the process and the technical requirements of the equipment allow it.

How long should a refrigeration system be measured?

There is no single correct measurement period for every facility. Measurements should cover representative load conditions: high and low production, defrost cycles, changes in weather and operation of different equipment combinations. A system with rapidly changing load also requires a suitably short data-logging interval.

Is free cooling always economical in industrial refrigeration?

No. It depends on the required fluid temperature, local climate, the number of hours with sufficiently low outdoor temperature, pump and fan power, and the cost of heat exchangers and controls. The potential must be calculated for the specific annual profile.

Why does a cold store use more energy despite similar production?

Possible causes include higher outdoor temperature or humidity, warmer incoming product, more frequent door openings, poorer heat-exchanger performance, changed setpoints, more defrost cycles or a different compressor sequence. A monthly kWh reading alone is therefore not enough for diagnosis.

Sources and technical references

• U.S. Department of Energy – Industrial Refrigeration, Better Buildings & Better Plants
• U.S. Department of Energy – Process Cooling and HVAC, Better Buildings & Better Plants
• UK Government Energy Technology List – Refrigeration Compressors
• Oak Ridge National Laboratory / U.S. DOE – review of frost and defrost technologies
• Regulation (EU) 2024/573 of the European Parliament and of the Council on fluorinated greenhouse gases