Commercial Ice Machine Hygiene: What Grows In There, and the Regime That Stops It

A food safety manager at a hotel group asked me to look at a machine that was working perfectly, which was the problem. Production was normal, the cubes were the right size, and every one of them carried a faint taste of the cellar it stood next to. When I lifted the curtain, the bin liner above the ice line had a soft grey film that came away on a glove, and the scoop was lying buried in the ice with its handle down.

The machine was not faulty. It was dirty, and dirty is a different job with different chemicals and a different owner.

Ice is legally a food and an ice machine is food production equipment. Cloudy, soft or odd-tasting ice is a water quality and hygiene signal rather than a refrigeration fault. The contamination usually does not arrive in the water, it grows inside the machine, which is why wiping out the bin is not cleaning. A regime that works separates three jobs: descaling, which is chemistry; sanitising, which kills what is living in there; and handling discipline, which stops your team re-contaminating clean ice with a scoop.

The law is short and it is unambiguous

Retained EU food hygiene law deals with ice in two sentences. Annex II, Chapter VII of Regulation (EC) No 852/2004 states: "Ice which comes into contact with food or which may contaminate food is to be made from potable water or, when used to chill whole fishery products, clean water. It is to be made, handled and stored under conditions that protect it from contamination."

Made, handled and stored. Potable water at the inlet satisfies the first word only.

Chapter V of the same Annex covers the equipment. It requires that all articles, fittings and equipment with which food comes into contact are to "be effectively cleaned and, where necessary, disinfected", and adds that "Cleaning and disinfection are to take place at a frequency sufficient to avoid any risk of contamination". Notice what is missing. There is no interval in the law. The frequency is yours to set and yours to justify, which makes it a HACCP question rather than a housekeeping one. Chapter V also requires equipment to "be installed in such a manner as to allow adequate cleaning of the equipment and the surrounding area", which a machine wedged into a cupboard at fit-out quietly fails.

The contamination is not in your water, it grows in the machine

There is decent evidence behind that. A 2025 study in Foods sampled ice and the water feeding it at the same premises, 108 paired samples taken from the storage compartments of machines in bars and restaurants in Apulia (Caggiano et al., Food Ice Hygienic Quality Investigation from Public and Collective Catering, Foods 2025, doi:10.3390/foods14071146).

The ice was consistently dirtier than the water that made it. Median total bacterial count at 22 °C was 175 CFU/mL in ice against 43 CFU/mL in water. The fungal result is the one worth reading twice: "The percentage of fungal 0-count samples was 8.33% for ice and 64.81% for water", so around nine in ten ice samples grew fungi while most water samples grew none. Their conclusion is one sentence and it is the whole argument for a cleaning specification: "Ice contamination does not appear to be directly related to the hygienic quality of water, but likely linked to the production, storage and maintenance of ice machines."

That is one Italian region, not a UK estate, and I would not pretend otherwise. But a ten-year global review reached the same place, reporting that "Almost all authors of the studies argued that ice contamination also depends on the hygienic–sanitary quality of the ice-making machines" (Triggiano et al., State of the Art in Hygienic Quality of Food Ice Worldwide, Microorganisms 2024, doi:10.3390/microorganisms12040690).

The mechanism is not mysterious. Inside a cube machine there is a permanently wet, cool, dark set of food contact surfaces that never freeze and never dry: the reservoir, the pump, the distributor, the underside of the curtain, the bin liner above the ice line, the bin drain. Air is drawn through the machine and the bin door opens forty times a shift in a kitchen full of airborne yeast and flour. Freezing does not sterilise anything. It preserves it.

The best UK data we have is about handling

The largest UK survey of catering ice I know of is twenty-five years old, which tells you how little attention this gets. Nichols, Gillespie and de Louvois examined 4,346 samples from retail and catering premises (The Microbiological Quality of Ice Used to Cool Drinks and Ready-to-Eat Food from Retail and Catering Premises in the United Kingdom, Journal of Food Protection 2000, doi:10.4315/0362-028X-63.1.78).

Most ice was acceptable. What is interesting is what happened to it after it left the machine. Ice on food displays was dirtier than ice for drinks, "with 23% containing coliforms, 5% E. coli, and 8% enterococci at 10(2) CFU/100 ml or more". And the line I quote to bar managers: "The microbiological quality of ice used to cool drinks was poorer when melt water was present in the ice buckets."

Standing melt water is the common factor between a bad ice bucket and a bad bin drain. Ice that has partly melted has been through a liquid phase in a warm open container, and that is where growth happens.

Cloudy, soft or fast-melting ice is a water reading

A commercial cube machine makes clear ice on purpose. Water flows slowly over a cold plate and freezes in layers, and because dissolved solids and air do not freeze at the same rate as water, they get pushed off the growing cube rather than trapped inside it. That is the whole trick. Interfere with it and the cube goes white in the middle.

Two things interfere. The first is what is in the water. Pentair, who make Everpure filtration for this application, put it plainly on their own ice machine filtration page: "Without filtration, these contaminants could result in hollow, cloudy, bad-tasting ice." The second is a fouled water distribution system that has stopped wetting the plate evenly.

So cloudy ice is rarely a refrigeration problem. It is telling you about the water and the state of the wetted parts, and it costs you at the bar, because a cube full of trapped solids and air melts faster and dilutes the drink sooner. If the ice has gone cloudy and the machine is slowing down or failing to release its harvest, you have moved out of hygiene and into scale and mechanics, which is a different sequence and is covered in our post on a commercial ice machine that has stopped making ice.

Descaling is not sanitising

This is the most common gap I find in an existing maintenance schedule. Somebody has bought a descale. Nobody has bought a sanitise. They are two different chemicals doing two different jobs, and the manufacturers set them out as two separate procedures for a reason.

Hoshizaki's cleaning sheet for its DCM range runs "Cleaning" with a scale remover and "Sanitizing" with diluted chlorine bleach as distinct sequences, each with its own soak and flush, and states: "The icemaker/dispenser must be cleaned & sanitized at least twice a year." (Hoshizaki cleaning sheet 51051, PDF.)

Manitowoc go further in a service bulletin on heavily scaled flaker and nugget machines: "To prevent this problem Manitowoc recommends a preventative maintenance cleaning/sanitizing every 30 days ... and a standard cleaning/sanitizing every 6 months". The same document notes that "Local water conditions may require treating the water supply before it enters the ice machine to prevent excessive mineral build-up", and describes the point of no return, where the machine is scaled enough that "you will be unable to run a standard cleaning procedure making ice with the cleaner and sanitizer" because it trips out before the cycle finishes (Manitowoc Service Communicator, Cleaning Heavily Scaled Flake/Chiplet/Nugget Ice Machines, PDF).

Two things to check in a contract. That the specification names both operations and both products in the manufacturer's order, rather than saying "clean and check". And that the engineer's report records what actually went in, because a descale on its own leaves the biology exactly where it was.

The parts nobody cleans

Chemistry through the water circuit deals with the wetted parts. It does nothing for the half of the machine your ice sits in, and that half is a site responsibility.

• The bin liner above the ice line. Never frozen, permanently damp, invisible unless somebody empties the bin to bare plastic. This is where I find film.
• The curtain or deflector and its hinge. Splashed constantly, dried never.
• The bin door seal and its underside, where humid kitchen air condenses every time the door opens.
• The bin drain and trap. A slow drain leaves standing melt water in the bottom of the bin, which is the ice bucket problem reproduced inside your machine.

Then the scoop, which causes more trouble than the rest combined. A scoop stored in the ice is a food utensil sitting unwashed in a food product for a week. It needs its own holder outside the bin and the same wash cycle as any other utensil. Glasses used as scoops are worse, because glass in a bin means dumping the bin, defrosting it and starting again.

Water treatment in a hard water area

In London and the South East this is not optional. Thames Water state that "Most of the water in the South-East of England is hard in nature", and classify hard water as 200 to 300 mg/l as calcium carbonate, with very hard above 300 mg/l (Thames Water, water hardness).

Two things follow. Filter life is a function of throughput, not the calendar. Hoshizaki's own spec for the H9320-51 cartridge is to "Replace approximately every 21,000 gallons or 6-12 Months depending on usage and water condition", so a hotel bar and a village pub cannot sit on the same schedule. Ask for changes driven by volume, with a date written on the filter head.

And a filter is not a disinfection device. Hoshizaki print the boundary on the same page: "Do not use where the water is microbiologically unsafe, or with waters of unknown quality without adequate disinfection" (Hoshizaki H9320-51 water filter system). One caution on softeners while you are at it. Thames Water say "we don't recommend using softened water for drinking or cooking" and advise a separate unsoftened tap. Since ice is a food, it is worth knowing which side of the building's softener your machine is plumbed into before a refit decides it for you.

Who owns what

Here is how I would scope it in a contract. The interval column deliberately invents no numbers, because the law gives none. It says what should decide the interval, and where a manufacturer has set a minimum, it names them.

Task	Owner	What sets the interval	Evidence
Scoop washed and stored in its own holder outside the bin	Site team	Your utensil washing rule. It is a food utensil	Cleaning sign-off, and a holder that is actually fitted
Bin door, handle and panels sanitised; check for standing water under and behind	Site team	Daily opening checks	Cleaning record
Bin emptied to bare liner, interior, curtain, deflector and door seal washed and sanitised	Site team, at a quiet trading point	Your HACCP plan, set to Chapter V's "frequency sufficient to avoid any risk of contamination"	Signed record naming the product and the date the bin was last emptied
Water filter change	Engineer or trained site staff	Throughput. Hoshizaki: "approximately every 21,000 gallons or 6-12 Months"	Dated label on the filter head
Descale the water circuit, evaporator, distributor, reservoir and pump	Engineer	Manufacturer minimum plus local water. Hoshizaki: "at least twice a year"	Service sheet naming the descaler
Sanitise the same circuit, as a separate operation	Engineer	Manufacturer minimum. Manitowoc: a PM clean/sanitise "every 30 days", standard "every 6 months"	Service sheet naming the sanitiser separately
Condenser clean, air clearance and ambient check; bin drain and trap cleared	Engineer	Every service visit	Service sheet with clearance recorded
Review whether the intervals are working	Food safety manager	Evidence from site, not the original assumption	HACCP review record

Putting ice into the HACCP plan properly

Article 5 of 852/2004 requires that "Food business operators shall put in place, implement and maintain a permanent procedure or procedures based on the HACCP principles", and the seventh of those principles is "establishing documents and records commensurate with the nature and size of the food business to demonstrate the effective application of the measures outlined in subparagraphs (a) to (f)" (Article 5, Regulation (EC) No 852/2004).

Read that against an ice machine and three things follow. The machine is a step in your process and should be named as one. The cleaning frequency you chose is a control measure that needs a reason behind it, not a number somebody inherited. And the records have to demonstrate it happened, which a service sheet reading "ice machine checked" does not.

The same Article adds: "When any modification is made in the product, process, or any step, food business operators shall review the procedure and make the necessary changes to it." Moving a machine, changing the filtration or doubling the covers all count. It is the same discipline behind proving your temperatures and the UK food storage temperature rules. Ice deserves the same treatment and gets less, because it carries the risk with none of the alarms.

Frequently asked questions

Why are my ice cubes cloudy or white in the middle?

Because something has interrupted the slow layered freeze that pushes dissolved solids and air off the cube. That is either the incoming water or the water distribution inside the machine. Pentair say that without filtration, contaminants "could result in hollow, cloudy, bad-tasting ice". Start with the filter and the water, then have the distributor and plate inspected.

Why does our ice smell or taste of the kitchen?

Ice takes up volatile compounds from the air in the bin and from the surfaces it touches, so it acquires the smell of whatever the machine is breathing. Common sources are an unsanitised bin, standing melt water in the drain, taste and odour compounds in an untreated supply, and food stored in the bin to chill it. Filtration handles the chlorine side, and Hoshizaki quote up to 97.3% chlorine reduction for their cartridge, "eliminating tastes and odors from the ice". The rest is cleaning.

Is the pink or black film in our ice machine mould?

It is a biofilm, a mixed community of bacteria and fungi rather than one organism, and colour is not a reliable way to identify it. What matters is the response. Take the machine out of service, dispose of the ice, empty the bin to bare liner, run a full clean and sanitise to the manufacturer's procedure, then work out why it grew: interval too long, bin never properly emptied, standing water, or a machine sited where it cannot be cleaned. The sampling work above recovered moulds from ice far more often than from the water feeding it, so treat it as a machine problem.

Can we just wipe the bin out instead of emptying it?

No, and it is the most common shortcut I see. Wiping reaches the surfaces you can get to with ice in the way, which are not the surfaces where growth sits. The bin has to come out to bare liner so the walls above the ice line, the curtain hinge, the door seal and the drain can all be washed and sanitised. Plan the ice cover for that window or it will keep getting deferred.

Our ice machine is leaking water. Is that a hygiene problem?

Treat it as one before you treat it as plumbing. The usual causes are a blocked bin drain, a drain run uphill or without an air gap, a split inlet hose, a cracked bin liner, or a door seal letting humid air in so the cabinet sweats. All of them put standing water where it should not be, and standing water at the base feeds the same growth you are trying to remove from inside. Clean accordingly, then fix the cause. Our guide to drainage and leak faults covers the same failure modes across refrigeration generally.

Does a water filter mean we can clean the machine less often?

It does not. A filter improves what goes in, which helps with scale, sediment, chlorine and clarity. It has no effect on what has already colonised the bin, the curtain and the drain. Hoshizaki print that boundary themselves. Filtration and sanitising are complementary, not alternatives.

What good looks like

The estates I never get emergency calls from have three things in common and none of them are expensive. The bin gets emptied to bare plastic on a defined day by a named role. The service report names two chemicals, not one. The filter has a date on it that somebody wrote by hand. The sites I do get called to have a machine nobody can reach behind, a scoop in the ice, and a schedule that says "clean and check". Nothing on that list is a refrigeration failure, which is exactly why it goes unnoticed until an inspector asks how the cleaning frequency was decided.

Ice machine hygiene is the easiest thing on a multi-site estate to leave to whoever is on shift, and the hardest to evidence afterwards. When you want it written down properly, talk to us about ice machine cleaning and we will go through the water, the intervals and the split between your team and ours before anything is committed to a schedule. It is worth looking at how that sits inside planned refrigeration servicing for the rest of your equipment, and where ice machine breakdowns are really coming from once the hygiene side is under control. Machine-specific codes are worth checking before anyone opens a panel, and we keep lists such as Polar ice machine faults for the makes we see most.

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: 22 September 2026