What a power event does to a refrigerator, weeks later

The calls we get after a bad storm season rarely arrive during the storm. They arrive two, three or six weeks later, from a household that has already put the event out of its mind and is now describing something that sounds unrelated: a cabinet that cools but no longer holds, a fan that runs when it should not, a display that resets itself overnight, an alarm that clears and comes back.

Roughly sixty-eight thunderstorm days a year land on this coast, and this stretch of Florida has had several multi-day outages inside a decade — Matthew in 2016, Irma in 2017, Helene in 2024. Both things matter, and they damage a refrigerator differently.

Surge is cumulative, not dramatic

A direct strike destroys a board outright and leaves obvious evidence. That is the rare case. Far more common is what sixty-odd storm days a year do over seasons: repeated small voltage transients on the incoming supply, each one well short of destructive, each one degrading the components that sit between the mains and the low-voltage electronics.

Nothing fails on the day. What changes is margin. A supply section that used to hold a clean rail begins to sag under load, and the first symptom is not a dead appliance but an inconsistent one — the machine behaves for a week, then does something stupid for an hour at three in the morning.

That is why intermittent complaints on this coast deserve to be treated as real rather than filed as user error. Intermittent is exactly what degraded, not destroyed, looks like.

An outage is two events, and the second one does the damage

Losing power is benign. A refrigerator that sits without power for four days is warm, unpleasant and full of spoiled food, but it is not broken.

The damage happens at both ends of the outage. Power on a failing grid rarely stops cleanly — it browns out, comes back, drops again, and the appliance is asked to start and stop repeatedly at reduced voltage. A compressor starting into low voltage draws hard, and the components that manage that start take the punishment.

Then the restoration itself: the whole neighbourhood’s motor load comes back within a few seconds of each other, and the surge on restoration is the moment most electronics fail. If you can get to the cabinet before the power returns, switching it off at the breaker and bringing it back deliberately an hour later is genuinely worth doing.

The other half of the restart problem is thermal. A cabinet that has been off for days comes back to a full warm load and a warm interior, and it runs flat out for hours to recover. A sealed system with a partly loaded condenser copes with that in winter and fails it in September, which is one of the reasons post-storm calls cluster.

Where variable-speed hardware changes the picture

Older refrigeration ran a fixed-speed compressor on a relay and a start device. Modern cabinets frequently use a variable-speed compressor driven by an electronic inverter, which is better at almost everything — quieter, steadier cabinet temperatures, less cycling — and which puts one more board between the mains and the motor.

That drive board is part of the sealed system’s control path, not a general-purpose control. When it degrades, the compressor may run at the wrong speed, refuse to start under load, or start and drop out after a few minutes, and the cabinet reads as a refrigerant fault to anyone diagnosing by symptom alone. Metering the electrical side before touching the sealed side is not caution for its own sake — it is the difference between a board and a compressor, and between two figures on an estimate that are nothing like each other.

What we check, and in what order

Supply first: voltage at the outlet under load, not at rest, and the state of the dedicated circuit the cabinet is meant to have. Then the low-voltage side, because a healthy control acting on a wrong reading is indistinguishable from a failed control until you meter it. Sensors are cheap, boards are not, and a sensor drifting a few degrees will produce every symptom a bad board produces.

Then the loads the board is switching — condenser fan, evaporator fan, defrost heater, damper — because a board that reads faulty is sometimes a board being asked to drive a shorted load. Only after all of that do we talk about replacing electronics.

Boards for these cabinets are ordered against the serial number rather than the model, they are not stocked locally in any depth, and they are rarely returnable once fitted. That is the practical reason we meter before we order, and it is also why we will tell you the wait before you commit rather than after.

What you can do

Keep the appliance on a dedicated circuit and know which breaker it is. If a named storm is coming and you have warning, chill the cabinet down a couple of degrees the day before, keep the doors shut, and switch the breaker off once the power drops. Bring it back deliberately after the neighbourhood has settled.

If something intermittent starts within a month or two of a power event, say so when you call. It changes what we meter first, and it usually shortens the visit.