Dry Plants Get Cleaned Wet and That Is the Hazard
The water goes into insulation and framing and stays there long after the floor looks dry
A low moisture plant pulled environmental swabs from the same blending room three weeks running and got the same organism back every time. The sanitation record for that room was complete and signed every night, and the crew had in fact done more than the written procedure asked of them, because a month earlier somebody had looked at the room, decided it was overdue, and run a hose through it from one end to the other.
That hose is the whole account, and the plant spent two months arriving at it.
What was believed, in order. First that the chemistry had weakened, so the sanitizer was changed. Then that the swabbing was picking up a sampling artifact, so the sites were moved. Then that a supplier had delivered contaminated material, so three lots were held. If you manage quality assurance or sanitation in a plant that runs a dry process, you will recognize the sequence, because it is the one the records support. Every one of those theories can be tested against a document. The actual cause had no document at all.
What the record showed, once somebody went back far enough, was a single line in a maintenance log about a deep clean of the blending room, dated eleven days before the first positive.
What the mechanism turned out to be is the part worth understanding, because it is not intuitive to anyone trained in a wet plant. In a dry process the controlling variable is water activity, which is a measure of how much of the water present is actually available to a cell rather than bound up in the product. Below a certain level of availability, bacteria do not grow. They do not necessarily die either. Many of the organisms that matter in low moisture production are tolerant of desiccation, and they sit in seams, in powder residue, on framing and inside equipment cavities in a state where they are doing nothing at all. A swab taken from that surface may well come back clean, because there is very little coming off it and nothing multiplying on it.
Introduce water and the arithmetic changes within hours. Availability rises, the dormant population becomes capable, and a room that was microbiologically quiet for a year starts behaving like a wet plant with no drainage and no air movement designed for it. The hose did not bring the organism into the blending room. It supplied the one thing the organism had been missing.
Get Daniel Lynn’s stories in your inbox
Join Medium for free to get updates from this writer.
The second half of the mechanism is where the time scale comes from. A dry room is not built to shed water. Floors are often flat, because nothing was ever meant to run across them. There are no trench drains in most of the room, and the ones that exist are sized for a spill rather than for a wash down. Water driven under pressure goes into places that have no way of giving it back: the gap behind an insulated wall panel, the inside of a hollow leg on a frame, the space above a false ceiling, the fiber of a cardboard pallet slip nobody removed. Pressure is what makes this different from a spill, because a spill sits on a surface and a jet finds every joint that was only ever meant to keep product out. The floor is dry by morning and looks finished. The structure is wet for weeks, and it is wet in exactly the places that are hardest to reach, which is the same geometry that produces a harborage point that keeps producing the same organism in a wet plant.
So the claim. A low moisture room needs a written dry cleaning procedure, its own dedicated tool set, and its own definition of done, and a plant that has none of those three is relying on the habits its crews brought from wet facilities. Those habits are good habits in the room they were learned in. Carried across, they are the hazard, because the method is matched to a process with different physics.
A working sanitarian will push back on that, and the pushback is reasonable. Dry cleaning does not remove everything. Rendered fat on a mixer shaft, a baked carbohydrate film on a dryer wall, a build up at a seal: there are soils that nothing short of water and chemistry will lift, and a plant that bans water absolutely will eventually be cleaning around a soil it cannot see past. There is also equipment that was designed to be cleaned in place with water and cannot be opened for a dry clean at all. The honest position is narrower than a ban. Water in a dry room should be an engineered event rather than an available option. It should be scheduled, confined to the equipment that needs it, carried out with the room opened up, and followed by a drying step with a stated end point. What should never happen is a hose in the hands of somebody who has decided a room looks tired.
The constraint that makes this hard is time, and it is specific. A wash down of a mid sized production room is a known quantity. A crew can scope it, and the visible water gives everyone a shared sense of progress. The equivalent dry clean takes longer, because vacuuming, brushing and scraping move soil at a slower rate than a pressure washer does, and because the only way to reach a frame cavity dry is to open it. It also produces airborne dust, which means respiratory protection and a dust control plan, and in some rooms a combustible dust assessment as well. Those hours have to come out of the same window production already resents giving up, and they come out of it every night rather than once a quarter, which is the part that loses the argument in a budget meeting.
And then there is the problem underneath all of it, which the field has not solved. A wet clean announces its own completion. The floor is clear, the water runs off clean, the room smells of the chemistry used. A dry clean has no such signal. There is no point at which a dry room visibly becomes finished, which leaves a supervisor judging the work by whether the crew looks like it worked hard. That is the same trap as a surface that looks clean and still fails a swab, except that in a dry room there is no equivalent of the visual cue at all, not even a misleading one.
I do not have a measurement to offer in its place. Water activity meters exist and are routine in laboratories for product, and I know of no plant measuring water activity at a floor seam or inside a wall cavity on a sanitation schedule, which is where the number would actually matter. Moisture meters read a surface and tell you very little about the volume behind it. Nobody has given the sanitation shift an instrument that answers the question the shift is being asked, which is whether the room is dry again. The nearest honest practice I know is a written sequence with a defined end state for each zone, a photograph of the finished condition, and a prohibition on introducing water that has a named person able to lift it. That is a procedural answer to a physical question, and it is weaker than I would like, in the same way that water standing inside a building resists every answer short of removing it.
What a competent team can do, in order. Write the dry procedure as a procedure rather than as a restriction, naming the tools, the sequence and the end state for each zone. Buy a second tool set in a different color and store it apart, because a brush that has been wet is a wet cleaning tool for the rest of its life. Make the introduction of water a change that requires authorization and a drying plan, the way a formulation change requires authorization. List the places in each room where water cannot be retrieved, and walk that list before any approved wet work. Put a monitoring site on each of those places and keep sampling them for a month afterwards, because that is the only evidence anyone is going to get. Then write the whole of it into the maintenance system as well, because the clean that caused this one was ordered by maintenance and never appeared in a sanitation document at all.
The plant in this account did all of that, eventually, and the room went quiet. What it never recovered was the two months spent testing the chemistry, the sampling plan and the supplier, none of which had failed. A dry room that has never been written down is cleaned by whatever the last crew learned somewhere else.
Daniel Lynn has worked in pathogen control and disinfection science since the late 1980s.
