Cold Room Not Cooling: The Diagnostic Sequence, In Order

A bakery group called us out to a walk-in freezer that had been drifting up overnight for weeks. The night shift had flagged it twice, and both times somebody had turned the setpoint down, which is the instinctive thing to do and precisely the wrong thing to do, because the room then iced its evaporator harder and held temperature worse than before. Nobody had gone near the refrigeration plant. The door closer had been unbolted months earlier because it kept catching the wheels of a rack trolley, and every night that door had been standing a finger's width off its own seal.

The plant was blameless. The room was not.

A cold room that will not hold temperature has one of two problems: too much heat going in, or not enough heat coming out. Diagnose it in that order. Door and door discipline first, condenser second, evaporator and defrost third, internal airflow fourth, controls and sensors fifth, refrigerant last. Refrigerant is the first thing most people say out loud and the last thing that should be tested, because everything above it is cheaper to rule out, faster to see with your own eyes, and more often the real cause.

Before you diagnose anything, decide what you are actually looking at

Three different situations get reported with the same sentence, and they need different responses.

The first is a room that is warm because it has been working. A goods-in delivery held the door open for twenty minutes and the room is pulling back down. That is not a fault, that is a cold room doing its job, and the only question worth asking is whether it recovers within a sensible window.

The second is a room that never reaches setpoint any more, but the plant is running. This is the common one and it is what the rest of this post is about.

The third is a reading that is wrong rather than a room that is warm. A displaced probe, a drifted sensor or a monitoring system reporting from the wrong point will all produce an alarm with nothing behind it. Before anyone is sent to site, put a calibrated thermometer in the room and compare it to the controller. That single check settles more overnight panics than any other, and it is the reason cold room temperature logging is worth having set up properly rather than as an afterthought.

Step one: the door, the seal and the habits around it

More cold room temperature complaints trace back to the door than to any component inside the plant, and the door is the only part of the system a site team can inspect without any risk at all.

Walk the perimeter of the closed door with the room light on and stand outside in the dark. If you can see a line of light, you have a leak path. Then check the things around it. Is the closer still fitted and still pulling the door shut against the gasket, or has it been removed because it was inconvenient? Have the hinges dropped so the door catches at the bottom corner and rests there? Has the strip curtain lost half its strips, or been tied back permanently because trolleys kept snagging? On a freezer room, is the heated frame actually warm to the touch, or has the door frozen itself into a permanently imperfect seal?

The frost tells you a lot. Frost concentrated around the door frame, ice building on the floor inside the threshold, or a rime line running along one edge of the gasket is a room being fed warm moist air continuously. That is a fabric problem, not a plant problem, and no amount of work on the condensing unit will fix it. Worn gaskets are a genuinely cheap repair and are covered in more detail in our guide to replacing a door gasket.

Then there is door discipline, which is a management question rather than an engineering one. A room propped open during a busy prep period, twice a day, every day, is a heat load that was never in the original calculation.

Step two: the condenser

If the door is sound, move to the heat rejection side. The condenser is where the heat the room loses is supposed to end up, and if it cannot get rid of that heat, the whole system loses capacity.

Look for three things. Fouling first: a coil packed with dust, grease from a nearby kitchen extract, leaf litter, or the fluff that comes off cardboard packaging. Fan failure second: a condenser fan that has stopped, or is running backwards after somebody rewired it, or is cycling on a faulty pressure switch. Siting third, and this is the one that ruins whole estates. A condenser in a tight yard, boxed in by a bin store, sitting under a kitchen extract discharge or facing a wall a foot away will recirculate its own hot air and effectively run in a much higher ambient than it was specified for.

Siting faults have a signature you can spot without any instruments. The room holds fine overnight and in the morning, then loses temperature through the afternoon and recovers again after close. If a room only fails in summer, and only in the second half of the day, stop looking inside the room. Fixing that properly is a design change rather than a repair, which is why we treat it as an installation question when it comes up on a survey.

Step three: the evaporator and the defrost cycle

Now go inside. The evaporator has to stay clear enough for air to pass through it. When it ices over, airflow collapses, the room warms, and the plant runs harder while doing less. This is the fault that looks most like a refrigerant problem and is not one.

An iced coil is a symptom. The question is why the defrost is not clearing it.

• The defrost is not initiating. A controller reset back to defaults, a schedule that was edited, or a timer that has failed will leave a coil to ice up over days.
• The defrost runs but does not terminate correctly. A failed termination sensor makes every defrost run to its time limit, which puts unnecessary heat into the room and can still leave ice in the corners of the block.
• A defrost heater element has gone open circuit. On a multi-element coil, losing one element frosts a section of the coil and leaves the rest clear, which is a distinctive pattern once you know to look for it.
• The drain is blocked or frozen. Meltwater that cannot leave refreezes on the tray and the coil, and on a freezer room a heated drain line or trap that has failed will do this within days. Standing water or an ice ridge under the evaporator is the giveaway.
• The defrost frequency no longer suits the room. A room whose use has changed, or which is now taking far more warm moist air through the door than it used to, needs a different defrost schedule to the one it was commissioned with.

The diagnostic tell for defrost faults is behaviour over time. If a manual defrost brings the room straight back to setpoint and it then degrades again over the following days, the refrigeration circuit is healthy and the defrost is the fault. If a manual defrost changes nothing, look elsewhere.

Step four: airflow inside the room

A cold room is designed around an air path. The evaporator throws cold air along the ceiling to the far wall, it drops through the product, and it returns to the coil. Block any part of that loop and you get a room with a cold corner, a warm corner, and an average that satisfies nobody.

What we find on site is almost always the same thing. Racking added after handover, stacked to the ceiling, sitting directly in the discharge. Pallets pushed up against the return face of the evaporator. Empty boxes stored on top of the coil because it was the only flat surface left. Product wrapped in pallet film so tightly that air cannot get between the cases.

This one matters commercially, because it is the failure mode that follows a business growing. The room was right when it was commissioned. Then trade doubled, stock levels went up, and nobody re-checked whether the air could still move. Overstocking is not a refrigeration fault, but it produces temperature excursions that look exactly like one, and it is the cheapest thing on this list to correct.

Step five: controls, sensors and the settings nobody logged

By this point you have ruled out most of the physical causes, so look at what the room has been told to do.

Check the setpoint and the differential against what was commissioned, not against what looks reasonable. Check where the air sensor physically sits, because a probe that has been knocked out of its pocket and is dangling in the discharge air will read cold and starve the room. Check whether any alarms have been muted or their thresholds widened, which happens quietly on sites where an alarm has been nuisance-tripping for months. Check whether the controller is sitting in a manual or hold state that somebody set during maintenance and never cleared.

Controller alarm codes are worth reading properly rather than clearing. Most cold room controllers distinguish between a probe fault, a high temperature alarm and a defrost fault, and knowing which one is showing narrows the diagnosis before anyone travels. We keep a set of manufacturer fault codes for the controllers we see most often in London.

Step six: refrigerant, and why it comes last

Low refrigerant charge does cause a room to lose temperature. It is genuinely on the list. It is last because it is the most expensive to investigate, requires a certified engineer, and is the diagnosis that most often turns out to be wrong when the room has an unbolted door closer or a coil full of grease.

It also carries legal weight. UK F-Gas guidance is explicit that a leak is repaired, not topped up. GOV.UK states: "If you find a leak during a check, you must repair it as soon as possible and repeat the test within a month to check the repair worked." (Checking F gas equipment for leaks, GOV.UK). The same guidance sets check frequencies by charge size in carbon dioxide equivalent, "at least once every 12 months" for equipment holding 5 to less than 50 tonnes of CO₂ equivalent and "at least once every 6 months" for 50 to less than 500 tonnes.

The person doing the work has to be qualified as an individual. GOV.UK is unambiguous: "You must have your own qualifications to work on equipment containing F gas, even if you work for someone else." (Qualifications to work with F gas, GOV.UK). If a contractor gasses a cold room without finding the leak and without giving you a record of what went in, you have paid for a fault to come back and you are carrying the compliance risk. Our guide to the F-Gas rules for businesses sets out what the operator is actually responsible for.

What a site team can check, and what needs a certified engineer

Check	Site team	Why
Door seal, closer, hinges, strip curtain	Yes	Visual and tactile only, no tools, no live parts
Calibrated thermometer against the controller reading	Yes	Separates a warm room from a wrong reading before anyone travels
Stock clear of the evaporator discharge and return	Yes	Housekeeping, and often the whole fault
Condenser coil visibly fouled or obstructed	Look, do not touch	Fins cut, fans start unexpectedly, and cleaning needs isolation
Photograph the controller display and any alarm code	Yes	Turns a phone call into a diagnosis
Changing setpoint or differential	No	Masks the fault and usually makes icing worse
Manual defrost, coil de-icing	No	Steam, hot water and screwdrivers all damage coils and cause leaks
Opening the electrical panel, testing components	No	Regulation 14 of the Electricity at Work Regulations 1989 restricts work on or near live conductors
Anything involving refrigerant	No	Certified persons only under F-Gas

That last row is not a formality. Regulation 14 of the Electricity at Work Regulations 1989 states: "No person shall be engaged in any work activity on or so near any live conductor (other than one suitably covered with insulating material so as to prevent danger) that danger may arise unless..." and then sets three conditions, including that it must be unreasonable for the conductor to be dead. A duty manager with a multimeter does not meet them.

The safety rules that apply while you are inside the room

Anyone who goes into a walk-in to look at a coil should know how they are getting out. The HSE guidance on chilled and frozen food production says: "Provide means of escape following entrapment inside walk-in refrigeration units, chill units and freezers." It adds: "Doors should be openable from the inside and lighting or other means provided to enable the door and opening device to be seen when the door is closed." (Chilled and frozen food products, HSE). The same page notes that a risk assessment may show trapped-person alarms are appropriate, which is worth reading alongside our post on walk-in safety alarms.

If you are inspecting a room during a fault, that is exactly the moment when the door mechanism may already be compromised. Check the internal release before you let it close behind you.

What this looks like across an estate

On a single site, this sequence is a checklist. Across ten or twenty sites, it is data. Ask your contractor to record which step resolved each cold room callout, and after a couple of quarters you will be able to see something a single visit never shows you.

The patterns are consistent. If most of your cold room jobs resolve at step one, you have a door and training issue, not a refrigeration issue, and the money is in door hardware and a short toolbox talk. If they resolve at step two and cluster in July and August, you have a condenser siting problem that was repeated across the estate by whoever fitted them. If they resolve at step three, your defrost schedules were commissioned for rooms that are used differently now. Each of those has a different fix and a different budget line, and none of them are visible if every job is written up as "cold room not cooling, attended, resolved".

That is the argument for scheduled refrigeration maintenance that produces useful paperwork rather than a signature on a sheet. A service visit that records door condition, condenser condition, defrost performance and controller settings gives you a trend. One that records "system checked, working correctly" gives you nothing to act on.

Frequently asked questions

Why is my cold room running constantly but not getting cold?

Continuous running with a warm room means the system has lost capacity or is fighting a load it was not sized for. In practice that is a fouled or badly sited condenser, an iced evaporator with a defrost fault, a door that is not sealing, or stock blocking the airflow. Refrigerant loss produces the same symptom, which is why it gets blamed, but it is the least likely of the five and the only one that needs a certified engineer to confirm.

Should I turn the setpoint down if the cold room is warm?

No. If the room is not reaching its current setpoint, a lower setpoint will not help it get there. On a chiller room it usually increases icing on the coil, which reduces airflow further and makes the situation worse. Leave the setting alone and record what it was, because a changed setpoint hides the evidence an engineer needs.

Why does my walk-in only lose temperature in the afternoon?

That timing points at the heat rejection side. A condenser in a confined yard, under an extract discharge or in direct sun will be working against a much higher local ambient in the afternoon than it was specified for, and it recovers once the outside air cools. Watch it across a full day before assuming the fault is inside the room.

Is ice on the evaporator normal in a cold room?

A light even frost that clears at each defrost is normal. A solid block of ice across the coil face, ice building only on one section, or an ice ridge in the drain tray are all faults. The useful test is whether a proper defrost restores the room. If it does and the problem returns over days, the defrost system is where the fault sits, not the refrigeration circuit.

Can our maintenance team clean the cold room condenser?

Look at it, photograph it, report it. Do not clean it. Condenser fins bend and cut, fans can start without warning on a call for cooling, and the unit needs proper isolation first. Coil cleaning belongs in a planned service visit where it is done safely and recorded.

How quickly should a cold room recover after a big delivery?

There is no single figure, because it depends on the room, the plant and how much warm product went in. What matters is that you know what normal looks like for each room. Logged temperature data over a few weeks gives you a recovery profile, and once you have that, a room taking noticeably longer than usual is an early warning rather than a surprise.

Working the sequence beats guessing at it

Almost every cold room we attend for a temperature complaint is mechanically capable of doing its job. What has changed is the door, the airflow, the condenser's environment, or the way the room is used. Working the list in order costs you nothing and rules out the cheap causes before anyone starts talking about compressors.

When a walk-in on your estate keeps coming back with the same complaint and nobody has written down which step actually fixed it last time, that is worth a conversation. Talk to us about cold room breakdowns and we will tell you honestly whether you are looking at a repair, a settings problem or a room that has outgrown its plant.

About the author: Ali Elm runs Be Cool Refrigeration & Air Conditioning, a family-run London firm that has been installing, servicing and repairing commercial refrigeration and air conditioning since 2004. His F-Gas certified team has delivered more than 2,000 projects across London and the South East.

Last updated: 4 September 2026