Chillers often reach the end of their economic life on refrigerant before wear: R22 and R123 are almost gone, and HFCs are being phased down. Mechanically, condenser tubes foul without good water treatment, evaporator tubes rupture if flow fails, hermetic motors burn out and contaminate the circuit, and drives and controls lose parts support early. Rising approach temperatures, refrigerant top-ups and high purge run time are the early signs. Low delta T looks like a chiller problem but starts in the building.

What it is and how it is built

Central cooling plant produces chilled water for a whole building or campus: one or more chillers, their primary pumps, the chilled water pipework in the plant room, and the controls that sequence them. It works with the heat rejection plant (cooling towers, dry coolers or air-cooled condensers, which have their own page) and the air handling and terminal units that deliver the cooling to spaces.

Chiller types
TypeWhere it is foundCharacter
Water-cooled centrifugalLarge offices, hospitals, campuses; 1960s onwardMost efficient at full load; surge at low load on older fixed-speed machines; longest life
Oil-free magnetic-bearing centrifugalRetrofits and new plant from the 2000sExcellent part-load efficiency; electronics-dependent; fewer mechanical wear parts
Water-cooled screwMid-size commercial and industrialRobust; good part-load with variable speed; oil management matters
Air-cooled screw or scrollSmaller offices, retail, apartments, where towers are unwantedNo cooling tower; condenser coils outdoors set the life
Absorption (lithium bromide)Sites with waste heat or cheap gas; hospitals, tri-generationCrystallisation and corrosion risks; vacuum must be maintained
Heat pump and heat recovery chillersElectrification projects from the 2010sProduce heating and cooling; replace boilers as well as chillers
ReciprocatingOlder plant, 1970s to 1990sMostly replaced; spares scarce

Distribution arrangements matter as much as the machines. Constant primary systems run fixed flow through each chiller. Primary-secondary systems decouple chiller flow from building flow through a bypass. Variable primary systems vary flow through the chillers themselves. Each has its own way of going wrong, and most chilled water problems that look like chiller problems start in the distribution system or at the coils.

Plant capacity is sized for a design load that may no longer exist. Tenancy changes, more efficient lighting, higher occupant densities, data rooms and longer hours all move the load, so a plant can be both old and wrongly sized.

Service life and what shortens it

Typical service lives
ComponentTypical lifeWhat moves it
Water-cooled centrifugal chiller20 to 30 yearsWater treatment; overhaul history; refrigerant
Water-cooled screw chiller20 to 25 yearsOil management; run hours
Air-cooled chiller15 to 20 yearsCondenser coil corrosion, especially near the coast
Absorption chiller20 to 25 yearsVacuum integrity; corrosion inhibitor
Primary pumps15 to 25 years; seals and bearings 5 to 10Run hours; alignment; water quality
Starters and variable speed drives10 to 20 yearsHeat and dust in the plant room; harmonics
Plant controls and chiller panels10 to 15 years before parts support endsManufacturer support
Chilled water pipework (steel)40 to 60 yearsWater treatment; external condensation and insulation

Industry service-life tables (ASHRAE's service life database and CIBSE Guide M) give median lives in this range; the spread between a well-run and a neglected plant is wider than the spread between machine types.

What shortens life:

  • Poor water treatment on the condenser water side, which fouls and corrodes tube bundles.
  • Low flow and low temperature conditions that freeze and rupture evaporator tubes.
  • Oversizing after load reductions, which makes chillers cycle on and off at low load.
  • Refrigerant leaks that are topped up rather than repaired, bringing in moisture and air.
  • Deferred overhauls past the manufacturer's run-hour intervals.
  • Hot, dusty plant rooms that age drives and starters.
  • Coastal air on air-cooled condensers.

How it fails

Refrigerant obsolescence

A chiller's economic life is often set by its refrigerant before its mechanical life. CFCs (R11, R12) went out of production in developed countries in 1996. HCFCs (R22, R123) are in the last stage of phase-out: R22 production and import ended in the United States in 2020 and servicing with it has been banned in Britain since 2015. HFCs (R134a, R410A, R407C) are being phased down under the Kigali Amendment, with Australia and the United States both stepping down to 15 per cent of baseline by 2036 and new high-GWP refrigerants restricted in new chillers. As a refrigerant becomes scarce, a leak or a compressor failure becomes a replacement decision.

Tube bundle fouling, corrosion and rupture

Condenser tubes carry tower water and foul with scale, biofilm and silt when water treatment lapses; fouling raises condensing pressure, power draw and wear. Corrosion and erosion thin the tubes, especially at the inlet ends. Evaporator tubes can freeze and rupture if water flow drops while the chiller is running, or during low-temperature operation; a rupture floods the refrigerant circuit with water and usually ends the machine.

Compressor and motor failure

Bearings wear, oil degrades and loses its lubricating properties, and oil migration leaves compressors running dry. In hermetic machines, where the refrigerant cools the motor, a winding failure burns the insulation and contaminates the whole refrigerant circuit with acid. Centrifugal compressors surge at low load: flow reverses briefly, with a characteristic noise, and repeated surge damages bearings and impellers.

Air and moisture in low-pressure machines

Low-pressure centrifugal chillers (R11, R123) run below atmospheric pressure on the evaporator side, so leaks draw air and moisture in rather than letting refrigerant out. A purge unit removes them; high purge run time means a leak, and moisture forms acids that attack the motor and bearings.

Low delta T

When chilled water returns from the building only a few degrees warmer than it left, chillers cannot load fully, so more chillers and pumps run to meet the same load. The cause is usually in the building: fouled or bypassing coils, failed control valves, three-way valves left from earlier systems, or bypasses open. The symptom appears at the plant and is often misread as a capacity shortfall.

Pump failure

Mechanical seals leak, bearings wear, couplings misalign, impellers erode and motors fail. Pumps running against throttled valves or at the wrong point on their curve wear faster. A pump failure without standby takes a chiller out of service.

Starters, drives and controls

Starters, variable speed drives and chiller control panels fail with heat and age, and their parts stop being made long before the chiller is worn out. A ten-year-old drive or a twenty-year-old control panel can strand a sound machine.

Air-cooled condenser coil corrosion

Aluminium fins on copper tubes corrode quickly in salt air, separating from the tubes and losing heat transfer; capacity falls on the hottest days, when it is most needed.

What to look for

In the plant room

  • Nameplates: manufacturer, model, serial, year and refrigerant. Year and refrigerant together fix most of the position.
  • Oil stains under compressors and at fittings; refrigerant leak detector alarms; the smell of oil.
  • Corrosion and condensation on chilled water pipework; damaged insulation.
  • Pump seals weeping; noisy bearings; vibration on base frames.
  • Number of chillers and pumps running against the load at the time.
  • Plant room heat, ventilation, housekeeping, and refrigerant detection and alarm equipment.

In operating data

  • Condenser approach (refrigerant condensing temperature minus leaving condenser water temperature) rising over time.
  • Evaporator approach rising.
  • Chilled water supply temperature drifting up on hot afternoons.
  • Low chilled water delta T across the building.
  • High purge run time on low-pressure machines.
  • Energy per unit of cooling rising.

In the building

  • Comfort complaints concentrated on hot afternoons.
  • More chillers running than the load warrants.
  • Temporary chillers hired in summer.

In the records

  • Service reports with recurring faults and deferred recommendations.
  • Refrigerant log: additions and leak tests.
  • Oil analysis, vibration analysis and eddy current tube test results.
  • Overhaul history against run hours.
  • Energy ratings and consumption trends.

Telling causes apart

Reading the signs
What you seeMost likely causeWhat confirms it
Condenser approach rising over monthsFouled or scaled condenser tubesTube inspection; water treatment records; approach falls after cleaning
Evaporator approach risingLow refrigerant charge or fouled evaporatorRefrigerant log; leak test; evaporator inspection
Refrigerant added at each serviceLeakRefrigerant log; leak test locates it
High purge run time on an R123 or R11 machineAir leak on the low-pressure sidePurge records; leak test under pressure
Chillers cannot load; more chillers and pumps running than the load needsLow delta T in the buildingSupply and return temperatures; coil and valve condition
Surging noise at light loadOversized centrifugal at low loadLoad profile; surge events in logs
Acidic oil or refrigerantMoisture or motor winding failureOil and refrigerant analysis
Capacity loss on hot days on an air-cooled unitCondenser coil corrosion or fouling, or hot air recirculatingCoil condition; plant screen or wall blocking airflow

How serious it is

Findings are graded by what they demand: safety (can hurt someone now), failure (failed and damaging the building), condition (distress present and progressing), lifecycle (renewal due on age and exposure), deadline (a regulator or ordinance sets the date), check (investigate before sizing works) and nil (cosmetic or stable).

Grading central cooling plant findings
FindingSituationWhat it means for timing
Ammonia or large refrigerant leak; plant room without required refrigerant detection or ventilation; pressure relief discharging into the roomsafetyIsolate and ventilate now; plant room safety measures before restart
Chiller failed with no standby capacity in a building that needs cooling (hospital, data, summer occupancy)failureRestore or hire temporary plant now
Refrigerant service restrictions in force for the refrigerant useddeadlinePlan replacement or conversion before the next major leak or repair
Rising approaches, recurring leaks, worn pumps, unsupported drives or controlsconditionRepair, retube or upgrade in the near term
Chiller approaching end of life on run hours and age with no current faultslifecyclePlan replacement, sized to current load
No tube testing, oil analysis or load study on record; capacity questioned after tenancy changescheckPlant condition and load assessment before sizing works
Normal wear in a maintained plantnilRoutine maintenance

Investigation and testing

A mechanical services engineer leads, with the chiller service contractor present. On large plant the investigation is planned around the cooling season so that machines can be opened.

  • Records review: service reports, refrigerant log, oil and vibration analysis, energy data, overhaul history.
  • Plant inspection of chillers, pumps, starters, drives, pipework and plant room safety equipment.
  • Eddy current testing of condenser and evaporator tubes to measure wall loss and find defects.
  • Oil and refrigerant analysis for acids, moisture, metals and contaminants.
  • Vibration analysis of compressors and pumps.
  • Performance testing: measured cooling output against power input across load.
  • Load assessment: trended building load against installed capacity, to size any replacement correctly.
  • Refrigerant position: current and forthcoming restrictions for the refrigerant in use.
  • Distribution check: delta T, valves, bypasses and coil condition before blaming the chiller.

The report should give remaining life for each machine and pump, the refrigerant position, whether to overhaul, retrofit or replace, and the correct size and configuration for any replacement.

Remedies and what they involve

  • Water treatment and tube cleaning to restore condenser performance.
  • Leak repair rather than topping up.
  • Overhaul: compressor rebuild, bearing and seal replacement, motor rewind, at run-hour intervals.
  • Retubing of bundles with significant wall loss.
  • Variable speed drive retrofit on centrifugal and screw machines for part-load efficiency.
  • Controls upgrade to supported chiller panels and plant sequencing.
  • Refrigerant conversion to a lower-GWP alternative where the manufacturer supports it for that machine.
  • Fix the distribution: replace failed valves, remove three-way valves and bypasses, clean coils, to cure low delta T.
  • Replacement sized to the current load: often fewer or smaller machines, with variable speed, in a configuration that allows staged replacement; heat recovery or heat pump chillers where the building is electrifying.

Interactions: chiller replacement involves rigging and access (often through roofs, façades or by crane), electrical supply works, temporary cooling, controls integration and sometimes heat rejection and pump replacement at the same time.

Cost shape

Assessment and testing are lump sums. Overhauls, retubing, drive retrofits and controls upgrades are lump sums per machine. Replacement is a lump-sum project including the machines, rigging and crane hire, pipework and electrical connection, controls integration, commissioning, and temporary cooling for the change-over, which can be a large share on occupied buildings.

Replacement and major overhaul are lifecycle; repairs and leak repair are repairs and maintenance; refrigerant-driven replacement and plant room safety measures carry code compliance elements. The usual shape where condition or load is uncertain is two-part: assessment first, then replacement or retrofit sized from it.

Standards and jurisdiction

  • Australia. Refrigerants are controlled under the Ozone Protection and Synthetic Greenhouse Gas Management Act 1989, with HCFC imports reduced to a small servicing allowance that ends by 2030, HFC imports phased down to 15 per cent of baseline by 2036, and refrigerant handling licensed through the ARC scheme. Refrigerating system safety follows AS/NZS 5149, and pressure equipment in-service inspection AS/NZS 3788. Offices of 1,000 m² and more need a current energy certificate under the Commercial Building Disclosure programme when sold or leased, and NABERS ratings drive lettability.
  • United Kingdom. The GB fluorinated gas regulations restrict refrigerants by global warming potential and require leak checks at intervals set by the CO₂-equivalent charge; R22 servicing has been banned since 2015. Refrigerating system safety follows BS EN 378; pressure systems need a written scheme of examination under the Pressure Systems Safety Regulations 2000. Air-conditioning systems over 12 kW need an energy inspection at least every five years, and let commercial buildings need an EPC of at least E under the minimum energy efficiency standards.
  • United States. Refrigerant management follows Clean Air Act Section 608; the AIM Act phases HFCs down to 15 per cent of baseline by 2036, and EPA rules limit the global warming potential of refrigerants in new chillers from 2025. ASHRAE 15 governs refrigerating system safety and ASHRAE 90.1 efficiency. Building performance standards, such as New York City's Local Law 97, set emissions limits that drive plant replacement on a legislated timetable.

What to ask the owner or agent for

  • Mechanical maintenance contract and two years of service reports with recommendations.
  • Refrigerant log, leak test records and refrigerant type for each machine.
  • Installation and overhaul dates; run hours.
  • Eddy current, oil and vibration test results.
  • Water treatment reports for the condenser water system.
  • Energy consumption data and energy ratings.
  • Comfort complaint log and records of temporary chiller hire.
  • Plant room refrigerant detection and alarm records.

Sources

Related references: the Capital Register and the MR Index.