Get in Touch with Koven Air Company
A centrifugal chiller is a water-cooled vapor-compression machine that removes heat from chilled water by circulating refrigerant through an evaporator, a high-speed centrifugal compressor, a condenser, and an expansion or metering device. Operating conditions are also affected by pumps, cooling towers, water temperatures, controls, sensors, and the system load.
Updated August 2026
Centrifugal chiller performance has two boundaries. AHRI ratings compare the chiller package under defined conditions; plant measurements show what the chiller, pumps, tower, and controls deliver together at the site.
Quick orientation: machine versus plant
- Inside the chiller: evaporator, centrifugal compressor, motor, condenser, expansion or metering device, and local controls.
- Outside the package: chilled-water pumps, condenser-water pumps, cooling tower, distribution piping, terminal or process load, sensors, and plant controls.
- Two different questions: “How efficient is this package at a rating point?” and “How much cooling does this plant deliver per unit of site electricity?”
- Core operating terms: load, lift, flow, speed, inlet guide-vane position, and the stable compressor operating envelope.
What a Centrifugal Chiller Actually Does
A centrifugal chiller transfers heat rather than “making cold.” Chilled water collects heat from a building or industrial process, refrigerant carries that heat through the machine, and condenser water usually moves it to a cooling tower. Its useful system boundary therefore extends beyond the chiller cabinet.
Air-Conditioning, Heating, and Refrigeration Institute liquid-chiller guidance describes chillers as part of larger systems that include pumps, towers, controls, distribution, and load-side equipment. That distinction prevents a common reporting error: comparing a package rating with a utility meter that also includes auxiliary equipment.
Three loops are involved in a typical water-cooled plant:
- Collect load heat — the chilled-water loop returns warmer water to the evaporator.
- Move heat through refrigerant — evaporation, compression, condensation, and expansion repeat inside the chiller.
- Reject heat outdoors — the condenser-water loop carries load heat plus compressor heat to the cooling tower.
Comfort cooling, data centers, hospitals, manufacturing, and process cooling can all use this architecture. Required leaving-water temperature, hourly load shape, heat-rejection conditions, redundancy plan, and service resources decide whether it is a good fit for a particular project.
How the Refrigerant Cycle and Main Components Work
The four main components perform distinct jobs: the evaporator absorbs heat, the centrifugal compressor raises refrigerant pressure, the condenser rejects heat, and the expansion device lowers pressure before the refrigerant returns to the evaporator. Water and refrigerant remain in separate circuits while exchanging heat across tube surfaces.
| Component | What enters | What changes | What to watch |
|---|---|---|---|
| Evaporator | Low-pressure refrigerant and returning chilled water | Refrigerant absorbs water-side heat and boils | Leaving-water temperature, flow, approach, and fouling |
| Centrifugal compressor | Low-pressure refrigerant vapor | Impeller adds velocity; diffuser converts velocity to pressure | Speed, inlet guide vanes, lift, current, vibration, and surge margin |
| Condenser | Hot compressed vapor and cooler condenser water | Refrigerant rejects heat and condenses to liquid | Entering-water temperature, approach, flow, scaling, and tower operation |
| Expansion device | High-pressure liquid refrigerant | Pressure falls before the evaporator | Refrigerant level, control response, and stable feed to the evaporator |
The compressor is called “dynamic” because it first gives the refrigerant kinetic energy. Rotating impeller blades accelerate the vapor; stationary passages then recover much of that velocity as pressure. By contrast, a screw compressor traps and reduces the volume of refrigerant between rotating elements. That mechanical difference shapes each compressor’s operating map and capacity-control behavior.
📐 Engineering note
Heat rejected at the condenser includes both the cooling load collected at the evaporator and the energy added by the compressor. Condenser-side measurements should therefore not be expected to equal evaporator load alone.
Load, Lift, Flow, and Surge: Four Terms That Should Stay Separate
Cooling load tells how much heat the plant must remove; lift describes the pressure or saturation-temperature difference the compressor must overcome. Refrigerant flow, shaft speed, and guide-vane position determine whether the compressor can meet that duty inside its stable operating envelope. No single load percentage can describe all four.
What is chiller lift?
Chiller lift is the compressor duty created by the difference between evaporating and condensing conditions. Warmer condenser water, a dirty condenser, lower chilled-water temperature, or another pressure-side constraint can raise lift even when the building load is unchanged. Higher lift generally asks the compressor to produce more pressure rise.
Variable speed can reduce impeller work at part load, but speed cannot fall without limit. AHRI’s technical overview notes that centrifugal capacity control commonly combines inlet guide vanes with speed control and that useful speed reduction depends on lift falling with the operating condition. Controls must still produce the head the refrigerant circuit requires.
What causes centrifugal-chiller surge?
Centrifugal-chiller surge is an unstable condition that can occur when the imposed head exceeds what the compressor can sustain at the given refrigerant flow, speed, and geometry. Low load may move an operating point toward that boundary, but low load alone is not proof of surge.
Purdue compressor-engineering conference research on variable-speed centrifugal chillers describes the relationship among the compressor map, surge line, lift, flow, speed, and variable geometry. Because its authors were affiliated with a chiller manufacturer, the paper is used here for the mechanism—not as proof of a universal savings percentage.
Surge Triangle: required head × flow × capability
- Required head: the pressure rise imposed by the evaporator and condenser conditions.
- Refrigerant flow: the flow associated with the current cooling duty and machine condition.
- Available capacity: the pressure rise the compressor can provide at the current speed and guide-vane position.
Credible assessment considers all three. Alarm text alone is not a compressor map.
How to Read Centrifugal Chiller Efficiency Without Mixing Boundaries
COP and kW/ton describe cooling output relative to input power, but every result needs a declared equipment boundary and operating condition. IPLV and NPLV support standardized package comparisons across part-load points; they do not reproduce a site’s hourly load, water temperatures, weather, fouling, staging, or auxiliary power.
Whole-plant energy consumption includes the chiller package plus the pumps, cooling tower, and other auxiliaries inside the declared meter boundary.
AHRI 550/590 and 551/591 apply standardized performance-rating methods to factory-made vapor-compression water-chilling packages. That common method is valuable for comparison. It becomes misleading only when a package rating is presented as if it were measured annual plant performance.
Do
- Name the rating or measurement condition.
- State which equipment is inside the power boundary.
- Synchronize cooling-output and power intervals.
- Compare like water temperatures and load conditions.
Don’t
- Treat IPLV as a site energy forecast.
- Compare chiller-only power with whole-plant power.
- Use controller load when flow or sensors are unverified.
- Claim savings without a baseline and boundary.
Worked plant-COP example
Suppose a validated water-side interval has a mass flow of 120 kg/s, entering water at 12°C, leaving water at 7°C, and water specific heat of 4.18 kJ/(kg·K). Cooling output is 120 × 4.18 × (12 − 7) = 2,508 kW. If synchronized power is 360 kW for the chiller, 50 kW for chilled-water pumps, 35 kW for condenser-water pumps, and 20 kW for tower auxiliaries, total plant power is 465 kW and plant COP is 2,508 ÷ 465 = 5.39.
This is an example calculation and not a Koven Air rating claim. Actual calculations require using the site fluid’s properties, calibrated flow and temperature measurements, aligned timestamps, and the project-specified meter boundary. Glycol concentration and change in temperature will alter the heat-capacity and density assumptions.
Key takeaway
An efficiency number is comparable only after the operating condition, cooling-output method, and electrical boundary are named.
Why Part-Load Operation Is a Plant-Control Problem
Part-load performance emerges from coordinated chiller staging, water flow, tower operation, and temperature setpoints. Reducing compressor power can be a false win if pumps or tower fans consume more, heat-exchanger conditions deteriorate, or a machine is pushed toward an unstable operating point. Plant-meter data settles the argument.
The U.S. Department of Energy’s systems approach to central-plant performance treats chillers, pumps, towers, controls, distribution, and loads as interacting parts. DOE’s process-cooling guidance also points to coordinated compressor staging, variable flow, variable-speed pumps and tower fans, and water-temperature resets that respond to load and ambient wet-bulb conditions.
“Adopting a system-level approach that considers the interaction and configuration of individual components is the most effective way to generate impactful energy savings.”
Three false improvements to watch for
- The compressor uses less power, but auxiliaries use more. Lowering the condenser-water target may reduce compressor lift while forcing tower fans and pumps to work harder.
- Flow falls, but the measurement becomes unreliable. Operating a flow meter below its useful range—or accepting a biased temperature sensor—can make calculated cooling output jump while the process itself has barely changed.
- A reset saves power briefly, then creates instability. Aggressive chilled-water or condenser-water resets can affect coils, dehumidification, process limits, cycling, or compressor operating margin.
Staging decisions should consider current load, expected load change, each running machine’s stable range, auxiliary power, start constraints, and redundancy. Rather than asking “How low can one chiller run?”, ask “Which combination meets the load with stable operation and the lowest measured plant input?”
Koven Air’s chiller efficiency comparison tool can facilitate an initial comparison, but site-specific load profile and measurement boundary must guide project decision-making.
Measurement-First Commissioning and Trending Checklist
Useful commissioning data must connect load, water conditions, compressor state, and auxiliary power on a common timeline. Before changing control logic, verify sensor identity, calibration, units, sampling intervals, and meter boundaries. Clean trend sets often resolve apparent equipment problems before a mechanical intervention begins.
| Trend group | Minimum evidence | Decision supported | Limitations / not sufficient alone |
|---|---|---|---|
| Chilled-water temperatures | Entering and leaving values from identified sensors | Cooling delta-T and setpoint tracking | Cannot prove load without credible flow |
| Chilled-water flow | Measured flow, meter range, and pump status | Water-side cooling calculation | Cannot prove accuracy without meter validation |
| Condenser-water temperatures | Entering and leaving values | Heat-rejection and lift context | Cannot isolate fouling from flow or tower effects |
| Condenser-water flow | Measured flow and pump state | Condenser heat-transfer check | Cannot define compressor duty without refrigerant conditions |
| Refrigerant conditions | Evaporator and condenser pressures or saturation temperatures | Lift and operating-envelope review | Cannot confirm flow or sensor calibration |
| Compressor state | Speed, guide-vane position, current, power, and demand limit | Capacity-control response | Cannot label surge without pressure and flow context |
| Chiller status | Enable, start, stop, load command, and alarm chronology | Cycling and sequencing review | Cannot explain cause without surrounding process data |
| Auxiliary equipment | Pump and tower status, speed, and power | Whole-plant energy balance | Cannot allocate unmetered shared loads exactly |
| Weather and load context | Outdoor wet-bulb, process state, and served-load status | Comparison of like operating periods | Cannot replace water-side or electrical measurement |
| Time alignment | Common time zone, synchronized stamps, and known sample handling | Valid interval-by-interval comparison | Cannot correct biased or mislabeled sensors |
An operating-event evidence map
- High plant kW per unit of cooling: confirm the meter boundary, then separate compressor, pump, and tower power before changing a setpoint.
- Low chilled-water delta-T: verify temperature bias and flow first, then inspect bypass flow, valve behavior, coil or process conditions, and sequencing.
- Repeated surge alarms: synchronize pressures, flow rate, compressor speed, guide-vane position, water temperature, and transition timing with each event.
- Frequent start/stop cycles: analyze load rate-of-change, minimum stable operation, deadband, sensor distortion, and start/stop sequence.
Troubleshooting should proceed systematically. Crosscheck data, establish boundary, pinpoint operating condition, and then proceed to mechanical and control changes. Erroneous sensors can guide service personnel to properly functioning equipment.
Centrifugal Versus Screw Chillers: Use the Load Profile, Not a Slogan
Centrifugal and screw compressors operate on different principles, so neither machine type can be assumed to be the best fit for every duty or a universal winner. Critical considerations include hourly load, water temperatures, heat-rejection conditions, available space, noise, refrigerant strategy, and maintenance planning.
| Decision factor | Question for a centrifugal option | Question for a screw option | Evidence to request |
|---|---|---|---|
| Load profile | How many hours fall inside the proposed stable map? | How does the selected configuration cover low and changing loads? | Hourly or sub-hourly load model |
| Water temperatures | What lift follows from the chilled- and condenser-water profiles? | What compression and control range follows from those profiles? | Design and off-design water-temperature schedule |
| Redundancy | Can the plant stage large units without leaving a poor operating point? | Would smaller modules better match the required failure tolerance? | N, N+1, or process-continuity requirement |
| Service model | Is trained support available for the selected compressor and controls? | Are parts, oil-management skill, and overhaul support available? | Service plan, parts list, and response commitment |
| Plant interaction | Can pumps and towers support the intended lift and flow range? | Can the same plant architecture support the proposed staging? | Sequence of operations and plant simulation |
This guide intentionally stops at decision criteria. Model specifications, project sizing, commercial terms, and quotations belong on Koven Air’s page for centrifugal chiller selection and specification options, where they can be evaluated against an actual duty profile.
When a Centrifugal Chiller Is—and Is Not—a Reasonable Candidate
A centrifugal chiller is a reasonable candidate when the cooling duty, water-cooled plant, load profile, operating temperatures, instrumentation, controls, and service plan support its stable operating range. Another architecture deserves equal review when the load is small or highly intermittent, heat rejection is impractical, or modular redundancy dominates.
Koven Air’s overview of industrial chiller system types provides the broader category context.
Reasonable conditions to investigate
- Sustained medium-to-large cooling duty
- Practical condenser-water and tower system
- Load and weather profiles compatible with a stable map
- Instrumentation and plant-level controls are part of the project
- Water-chemistry management and qualified service are available
Conditions that require a broader comparison
- Very small, intermittent, or sharply cycling load
- No practical water-side heat rejection
- Frequent duty outside the proposed stable envelope
- Project needs favor smaller independent modules
- Local service or water-quality controls are not ready
Before asking for a model, prepare the design load, hourly load profile, leaving chilled-water requirement, condenser-water profile, ambient design data, electrical constraints, redundancy basis, water-quality plan, controls interface, and measurement boundary. Koven Air’s industrial chiller selection finder can organize the first pass; final sizing still belongs in a project-specific engineering review.
Frequently Asked Questions
What are the main components of a centrifugal chiller?
Centrifugal chiller packages center on an evaporator, centrifugal compressor, condenser, and expansion or metering device, supported by a motor and local machine controls for stable operation.
Inside the evaporator, heat passes from chilled water into the refrigerant. Compression raises vapor pressure; condensation transfers heat into condenser water; expansion drops refrigerant pressure before the next pass. Motor and local controls support the compressor. Pumps and the cooling tower normally sit outside the basic refrigerant cycle, within the wider plant boundary.
What is a centrifugal chiller used for?
A centrifugal chiller supplies chilled water for larger comfort-cooling and industrial process-cooling loads when a central water-cooled plant matches the project duty and operating conditions.
Central plants can serve offices, hospitals, campuses, data centers, factories, and controlled processes. Application fit is not determined by the building label. Engineers compare the hourly load shape, leaving-water requirement, condenser-water and weather profile, redundancy, plant-control capability, water-chemistry management, service access, and the proposed compressor operating envelope.
What causes a centrifugal chiller to surge?
Surge can occur when the required compressor head exceeds stable capability at the current flow, speed, and guide-vane position, pushing operation beyond the stable compressor boundary.
Low load may contribute, yet it is not sufficient proof. Align evaporator and condenser pressures or saturation temperatures with credible flow, compressor speed, guide-vane position, current or power, chilled- and condenser-water temperatures, and the alarm timeline. That record shows whether required head, refrigerant flow, and available compressor capability crossed the stable operating boundary.
What is the difference between centrifugal and screw chillers?
A centrifugal compressor adds velocity and converts it to pressure, while a screw compressor reduces trapped refrigerant volume, giving each machine a different operating map and capacity-control behavior.
That mechanical difference changes each machine’s operating map and capacity control. Selection still needs a project load and water-temperature profile.
How long can a centrifugal chiller last?
No single lifespan applies to every centrifugal chiller; operating hours, lift, cycling, water quality, maintenance, controls, and overhaul strategy govern service life at each site.
Lifecycle planning reviews tube condition, refrigerant-circuit integrity, bearings or magnetic-bearing systems, motor and drive condition, control support, water-chemistry management, and parts availability. Repeated high-lift operation or poor water conditions may change the maintenance path. Request an inspection-based overhaul plan rather than accepting an unsupported universal service-life number.
Which data should be trended before changing chiller controls?
Trend both water loops, refrigerant conditions, compressor state, auxiliary equipment, load context, weather, setpoints, and alarms on one synchronized timeline before changing any control sequence.
Start with entering and leaving water temperatures, credible flow, evaporator and condenser conditions, speed, guide-vane position, chiller power, auxiliary status, weather, and aligned alarms. Validate sensors before drawing conclusions.
Practical Takeaway
Plant-level assessment treats a centrifugal chiller as a refrigerant machine working inside a chilled-water system. Standard ratings help compare packages; synchronized measurements explain site operation. Keeping load, lift, flow, surge, and the meter boundary separate produces better control decisions and a more defensible equipment selection.
Move from guide to project requirements
Consult a centrifugal-chiller expert to look at your load profile, water temperatures, redundancy, controls, and measurement boundary.
For company context, see Koven Air’s HVAC manufacturing background.
Research and scope notes
This article separates standardized centrifugal-chiller ratings from plant-level measurement and uses public AHRI, U.S. Department of Energy, and Purdue technical material for the cycle, controls, and operating-envelope explanations. Product specifications and commercial claims are intentionally left to the linked solution page.
References & Sources
- Liquid Chillers — Air-Conditioning, Heating, and Refrigeration Institute
- Variable-Speed Centrifugal Chiller Control for Variable Primary Flow Applications — Purdue University e-Pubs
- AHRI 550/590 and 551/591 Performance Rating Standards — Air-Conditioning, Heating, and Refrigeration Institute
- Systems Approach to Central Plant HVAC Performance — U.S. Department of Energy Better Buildings
- Process Cooling and HVAC — U.S. Department of Energy Better Plants
![Commercial Rooftop Unit Guide: Types, Sizing & Costs [2026]](https://kovenair.com/wp-content/uploads/2026/07/commercial-rooftop-unit-guide-featured-2-150x150.png)