When a chiller alarms, the first reaction is often to press reset. That clears the evidence and may restart a machine with a closed valve, a dry evaporator or a bad sensor. The first 30 minutes should be spent capturing the operating state and separating a water problem from a refrigerant or electrical problem.
This article is a triage sequence, not a code list. The code tells you which protection operated; the snapshot tells you why.
Do not reset first
Before touching the panel, record the time, the alarm text, the operating mode and the last 10 minutes of data: chilled-water entering and leaving temperature, condenser entering and leaving temperature, suction and discharge pressure, compressor current, oil pressure and water flow status. Take a photograph of the screen. If the controller has a history buffer, download it before resetting.
A reset that succeeds does not prove the fault is gone. It may have been a transient flow switch or a sensor spike. A reset that fails repeatedly is telling you the protection is doing its job. If the controller has no history buffer, photograph the gauge panel every few minutes until the machine stops.
Five alarm families
| Family | Typical evidence | First check | Do not do |
|---|---|---|---|
| Water flow | Flow switch open, pump status off, pressure difference low | Pump, strainer, valve position, air in the loop | Bypass the flow switch |
| Pressure | High discharge or low suction | Condenser water, tower, refrigerant charge, sensor | Add gas before checking superheat |
| Electrical | Overcurrent, phase loss, motor overtemperature | Voltage, phase imbalance, insulation, contactor | Reset repeatedly without a megger test |
| Sensor | Out-of-range reading, fixed value, communication loss | Wiring, plug, calibration, replacement sensor | Replace the controller first |
| Compressor | Oil pressure, discharge temperature, liquid slugging | Oil level, oil filter, superheat, liquid line | Run at full load to see if it clears |
0–5 minutes: capture and isolate
Confirm whether the compressor is stopped and the pumps are running. If the evaporator is at risk of freezing, keep the chilled-water pump running and close the isolation valve only after the water is above the freeze point. Do not stop the pump on a low-temperature alarm; that removes the heat that is protecting the evaporator.
Check the water flow at the pump and at the chiller. A closed valve, a blocked strainer or an air lock can trip the flow switch. If the flow is low, fix the hydraulics before resetting. On a condenser high-pressure trip, check the tower fan, the condenser water pump and the approach; a tower that lost a fan can raise the condensing temperature within minutes.
5–15 minutes: separate water from refrigerant
If the water side is sound, look at the refrigerant side. A low suction pressure with a normal water temperature may be a charge issue, a restriction or a sensor error. Check superheat and subcooling; a low superheat with a low suction pressure suggests overcharge or a liquid restriction, not undercharge. A high discharge pressure with a normal condenser water temperature suggests a fouled condenser or a non-condensable gas.
For an electrical trip, check the voltage and phase balance before resetting. A single-phasing event can damage the motor; a megger test and a visual check of the contactor are cheap compared with a rewind.
15–30 minutes: decide
If the fault is a transient sensor or a flow switch that cleared, a supervised reset is reasonable. If the same protection trips twice within 30 minutes, isolate the machine and call for service. Repeated resets can turn a sensor fault into a compressor failure.
Write a short log: time, alarm, operating point, action and result. The log is more useful than a verbal handover, and it becomes the baseline for the next alarm. If the machine is under a service contract, send the log and the downloaded history, not only the alarm code.
What the code does not tell you
A high-pressure code does not tell you whether the tower fan is off, the condenser is fouled or the charge is high. A low-pressure code does not tell you whether the filter is blocked or the sensor has drifted. The code is the protection that operated; the cause is in the operating data. Treat the first 30 minutes as evidence collection, not repair.
The handover
At the end of the 30 minutes, write a handover that a second technician can use. It should state the alarm, the operating point, what was checked, what was ruled out and what is still unknown. A note that says “high pressure, reset OK” is not a handover; it invites the same fault to recur on the next shift.
If the machine is isolated, tag the isolator and record the time. If it is running, set a shorter alarm review interval and watch the trend for the next two hours. The first 30 minutes do not have to solve the fault. They have to prevent a second, more expensive fault while the cause is still unknown. A photograph of the gauge panel and a copy of the alarm log are part of that handover.
Common questions
Should I reset a chiller alarm immediately?
No. Capture the alarm text and the operating snapshot first. A reset clears the evidence and may restart the machine into the same fault. If the protection trips twice, isolate it.
What data should I record?
Time, alarm, chilled-water and condenser temperatures, suction and discharge pressures, compressor current, oil pressure, flow status and the last 10 minutes of trend data.
The flow alarm cleared after reset. Is it safe?
Only if you can explain why. A transient air lock or a pump start sequence may be harmless; a failing flow switch or a closing valve will trip again. Watch the next start.
Can I bypass the flow switch to keep the plant running?
No. The flow switch protects the evaporator from freezing. Bypassing it can destroy the heat exchanger. Fix the pump, strainer or air problem instead.
A high-pressure alarm trips at night. What is the first check?
Check the cooling tower fan and the condenser-water pump first. A lost fan or pump raises the condensing temperature quickly. Then check the condenser approach and the refrigerant charge.
How do I know if the sensor is wrong?
Compare the sensor with a calibrated thermometer or gauge at the same point. A fixed value, an out-of-range value or a reading that does not change with load is suspect.
What snapshot first?
Time, load, pressures, temperatures, flows, current, alarms and the last action. Photograph the screen. The snapshot is the evidence for every later decision.
When do I isolate?
On a leak, a phase loss, a loud noise or a repeated trip. Isolation protects the plant and the people. Do not reset a fault you cannot explain.
Further technical reading
For the code list, see Water Chiller Fault Codes and Troubleshooting. For high-pressure causes, see Chiller High-Pressure Alarm. For low-pressure, see Chiller Low-Pressure Alarm. For remote alarms, see Chiller Remote Monitoring.
Related reading
Chiller plant sensors · Chiller plant group control · Evaporator cleaning