Why a floor joint decides the job, not the tile
The surface of a quarry tile or a porcelain tile is close to impervious. The joint is the opposite: cementitious grout is porous, it is deliberately lower than the shoulder of the tile, and it is full of tiny channels. Anything spilled in a commercial kitchen that reaches the floor - oil from a fryer bay, fat from a grill line, sauce from a plating pass - migrates into those channels and sets there. Dust and food solids get trapped behind it. The result is a joint that feels rough to a fingertip and reads grey or brown even after the tile beside it has been wiped.
That geometry explains why the ordinary routine does not move the colour. A mop with a neutral cleaner brushes the crown of the joint and leaves the body of it alone. A scrubbing machine with a pad works on the tile face and glances off the recess. Spray and wipe reaches the tile, not the seam. So the first thing to confirm before choosing any method is not which machine to buy; it is what is actually in the joint and how deep the deposit goes. If the answer is set fat with trapped solids, the confirmation list is about heat, dwell and lifting. If the answer is mineral or lime deposit, or the joint itself is crumbling, steam is not the tool and the discussion changes completely.
The three separate confirmations people usually skip
There are three questions that must be confirmed before the first metre of joint is treated, and they are usually conflated, which is why deep cleans go wrong.
First, confirm that the joint can take work at all. A joint that is sound, a joint that has been washed out, a joint that is cracked, and a joint that was patched with a different material four years ago are four different situations. On the patched and washed-out groups the right answer is repair first, cleaning second, and the repair belongs to a flooring trade rather than a cleaning crew.
Second, confirm that the method suits this kitchen, not kitchens in general. The same joint construction behaves differently in a pastry section that runs dry and cool and in a fry line that runs greasy and warm, and the drainage arrangement under the units decides whether a wet method is even allowed here.
Third, confirm what will count as finished. A deep clean that has no written acceptance standard is judged by opinion the following morning, and the crew that did the work always loses that argument. Get the standard agreed before the first metre is treated.
The joint survey: what to write down, group by group
Do not survey the floor as one area. A kitchen floor divides naturally into groups that differ in soil, in access and in how much disruption the work causes: the line in front of the hot range and fryers, the plating and pass area, the servery or scullery, the walk-in chiller thresholds, the dry stores and corridor runs, and the areas behind and under low equipment where the routine mop never reaches. Survey each group separately and give each one three columns.
Column one is what the deposit is: loose dry crumbs and flour; soft fatty film; set fat with trapped solids; sticky sugar from drinks and syrups; or a mineral haze. Column two is what the joint is doing: sound and slightly stained; stained and rough; washed out or recessed; cracked; missing in runs; patched with a harder material than the surrounding line; or etched and sandy at the surface. Column three is how reachable it is: open floor with no obstruction; floor under wall-hung units where a lance can reach but a person cannot; a threshold with a change in level; a channel or a grating; a tight internal corner; the strip under a plinth.
The reachability column decides most of the decision. A joint that can be approached squarely with a wand tool and worked at walking pace is one job. The same joint under a counter, worked on hands and knees with one hand, is a slower, less repeatable job and needs a different tool and a different time allowance. Record this before anyone quotes you a method.
Two cheap checks belong in the survey and need no laboratory. Run a fingertip across a joint and then across a tile: if the joint is high or ragged rather than recessed, the deposit is being held by the joint profile, not only by its porosity, and mechanical work matters more than chemistry. Then press a damp white cloth along a metre of joint and look at it sideways; the colour it picks up tells you whether the film is fatty, proteinaceous or mineral, and mineral haze reads very differently from kitchen fat.
Condition before method: when steam is off the table
The most expensive mistake in a kitchen floor programme is choosing the method before recording the condition. Some joint conditions should stop the conversation about heated steam until a flooring trade has looked:
Joints that crumble to powder when scraped, or that have washed out so the tile edges are unsupported, need re-grouting, not cleaning; work on them now and the tiles begin to lift later. Joints that have been patched with a hard, rapid-set repair material take water and heat differently from the original line, so a single treatment across a mixed line will leave one section clean and the next section damaged. Joints that have been sealed or coated within the last few years should be tested on the smallest hidden patch you can find, because a worn coating can be removed in patches and re-coating a whole floor is a bigger project than the clean that triggered it. Floors with a failed waterbar or a cracked screed under the joint line should be treated as a building defect first; driving moisture into a floor that cannot dry is how a cleaning job becomes a structural one.
Write these exclusions down with a replacement method for each one and the trade-off that comes with it. Cold water and a brush for a joint that cannot take heat; a low-moisture foam and vacuum for a floor above a occupied space; hand work with a detailing tool where the machine tool will not fit; and repair by the flooring trade where the joint has failed. An exclusion list with alternatives is what makes a floor programme survive its first audit. An exclusion list without alternatives just moves the dirt somewhere else.
Where the water goes, and what has to keep running
In a kitchen, drainage and downtime decide feasibility more often than cleaning power does, and both belong in the confirmations before any equipment is specified. Before the first pass, get four answers on paper.
Where does the water for this work connect to, and can a hose reach the far end of the floor from one position? What drain does each area group discharge into, and is that drain allowed to receive grease-laden wash-off, or does the run-off have to be collected and removed as waste? Does the floor fall correctly toward the channels, and where does water stand overnight now? Then the operational question: how long can each group be closed, and who signs off closing the pass, the servery or a walk-in? A joint clean that leaves a kitchen closed past the start of service is a failed joint clean regardless of how good the joints look.
Drying is part of the same answer, not an afterthought. A kitchen floor must be dry enough to walk and to put trolleys back on before service. That means the sequence has to end with a deliberate drying step, with an extraction or mop lift immediately after rinsing, with doors or windows available for airflow, and with a realistic time allowance between the last wet pass and re-opening. Confirm the drying window for each group in the plan, not after it.
Electricity at the point of work is the fourth answer. Ask what socket is near the group being worked, whether the cable can be routed away from a wet floor without crossing a service aisle, and what the site policy is on cable protection in a wet area. On a kitchen floor with washdown every night, an unplanned trailing cable is both a hazard and a reason the job gets abandoned halfway.
How heated steam works on a joint, in plain terms
Steam is not a stronger version of a mop, and the phrase high-temperature should be checked against the equipment's own record before it is repeated in a specification. In practice there are two ways of working, and they differ in how much water reaches the floor, not in how hard they attack the soil.
The first is a low-moisture, dwell-led pass: heat is applied, left for a stated contact time so that the set fat softens, and the deposit is then lifted mechanically and the residue removed. Little water reaches the build-up, so drying is quick and the floor can be re-opened sooner. The second is a higher-volume wet pass or rinse, which carries more water into the joint and into the room, needs a drain or a collection route, and needs a longer drying window. Neither is "stronger". The choice is made per area group against the deposit and the room's drainage and closure limits.
Two habits cause most failed joint cleans. Skipping the dry stage - vacuum or sweep the loose soil before any heated pass - multiplies the water and turns soil into paste. Skipping the lift after the dwell leaves softened fat suspended, which re-deposits overnight and reads grey by the following morning with people saying nothing was achieved. Sequence, contact time and the lift matter more than the machine behind them.
If heated steam equipment is under consideration for this floor, ask the supplier for the figures that actually govern a joint: the temperature and pressure at the point where the tool meets the surface rather than at the generator, the steam volume and how it is selected between low-moisture and wetter working, how the unit is fed with water and from where, and how it is moved and stored in a kitchen with tight aisles and closed doors. Also establish, in writing and for the exact unit quoted, which of the range offers a dry-steam and wet-steam selection - the company's Fixed Steam Cleaning Machine PowerDe HW02 page is the record to ask about, and what its page states describes that machine and does not transfer to another model. How the unit takes its water supply, and from which point in the kitchen, belongs in the same written list. Those are row labels to be answered with the quotation, not assumptions to be read across a catalogue.
A bounded trial: the confirmation that settles the argument
Before committing a whole kitchen to anything, mark off one length of joint in a low-traffic group with tape and treat only that. Photograph it at an angle, as it stands, with a coin or a ruler in frame for scale. Write the trial conditions down while the equipment is cold - what was applied, for how long, how many passes, what was used to lift, how the residue was removed, how long the floor took to dry. Then photograph the same patch again at the same angle once it is dry, and inspect it a full working day later, once the kitchen has been back in service, because a joint that looks fine at seven in the evening and greys by midday has not been cleaned, it has only been moved.
Write the stop conditions before the first pass, not during it: joint material starting to soften or wash out, tile edges lifting, discolouration spreading beyond the treated length, a coating lifting, or water standing longer than the agreed drying window. Stop conditions written in advance are the only thing that makes a trial evidence rather than an improvisation.
Take the time and cost figures from the second run, not the first. The first run includes finding the reach, moving a trolley twice and discovering the obstruction. Two measured runs of the same taped length, timed, are the only defensible basis for pricing the floor; a single optimistic pass becomes the budget and then the argument.
What to ask the supplier in writing, and what to do with the answers
Cleaning equipment for food-service floors is rarely a catalogue purchase, and the company's own customized steam cleaning equipment enquiry page is where that conversation starts. Whatever route you take, ask the same questions of every supplier and hold the answers side by side, and ask for them for the exact unit quoted:
Whether this specific machine is offered with both low-moisture and wetter working, and how the operator switches between them. What the supplier advises for a tiled floor with grout joints in a commercial kitchen, stated as an application answer rather than a general capability claim. Which protective functions that unit has, listed by name without assuming a feature carries over from a page or a variant. How the unit is fed and where it can be plugged in near the work, plus what the connection looks like against the kitchen's taps. Whether any extraction or recovery function exists at all on the unit quoted, since it is not present across every model. Which accessories reach a recessed joint, a threshold, a channel and a corner, and which are supplied rather than optional. What the supplier's own exclusion list is for heated steam, and what they propose instead where the answer is no.
Keep the answers with the trial record. Where a supplier's stated application for a model is not supported by the trial on your own floor, the trial governs, not the datasheet. And keep the machine data separate from the cleaning result: figures describe a machine, while the result describes your floor.
Choosing the method row by row, not floor by floor
With the survey, the condition list and the drainage answers written down, the method becomes a set of small per-group choices rather than one big decision. A useful rule: pick the option that puts the least water into a group while still giving enough heat and contact time to soften the deposit, and pick the wetter option only where the group has the drainage and the closure window to accept it.
Behind and under low equipment, where reach is poor and a person cannot get close, a low-moisture pass with a detailing tool and a manual lift is usually the only repeatable answer. Along the hot line, where the deposit is set fat and thick, heat plus contact time plus a mechanical lift is the combination that changes the joint rather than the surface above it. In dry stores and corridor runs, sweep and vacuum plus a cold-water brush on a shorter cycle usually beats any heated work. At thresholds, channels and gratings, the constraint is the drainage and the loose debris in the channel, not the machine.
Whatever mix comes out of that decision, the whole floor is only as clean as the last rinse and the last drying step. A crew that treats joints and leaves residue and moisture in place will see the colour return within days and will be told the method does not work, which closes off better options for the next cycle.
Keeping the joints clean after the deep clean
A deep clean restores joints; the routine decides how long they stay that way. Two changes carry most of the benefit. First, deal with spillage as it happens at the source: a mat or drip tray under the fryer bay and the plating pass, and a wipe-down of the floor immediately under the line at the end of each service, before fat and solids walk into the joint. Second, put the joint in the routine rather than only in the deep clean: a scheduled vacuum or sweep along the joint line, and a low-moisture brush pass on the groups that take the most soil, on a cycle the kitchen can actually keep.
Set the interval from your own floor, not from a brochure. The trial measured how long the groups that were treated took to look dirty again after a production shift; use that observation to set the review cycle for each group, then re-inspect on the agreed date. A joint that is re-inspected on a date will be caught when it needs a brush pass, not when it needs another deep clean.
Who owns it, how often, and what happens when it fails
Write the outcome as one page per area group, with four entries on each line: the owner, the frequency, the method, and the defect route. Owner means a named role, not a department. Frequency comes from your trial and your re-inspection, not from a supplier's default. Method is written in operator order and ends with a dry floor. The defect route says who is told when a joint starts to wash out, crack or lift a tile edge - because that is a repair, and a cleaning crew should not be expected to carry it.
Give each page a review date and keep the trial photographs with it. When the next cycle is scoped, the pair of angled photographs is what tells you whether the interval was right and whether the exclusions still hold. For the wider question of which equipment classes are even in scope for a food-service floor, the company's PowerDe cleaning equipment range is the starting point to compare against your own group list and trial result.
A short confirmation list before the work starts
If you want the confirmations in one place: what the deposit actually is, tested by fingertip and cloth. What the joint is doing, recorded group by group, with failed and patched lines excluded and referred. Which groups can be closed and for how long. Which drain or collection route each group discharges to, and whether grease-laden run-off is allowed there. Whether the floor falls and dries correctly today. What power and water are available at the point of work and how cables stay out of wet areas. Which one taped length will be trialled, with what written conditions, what stop conditions and what acceptance photographs. And which written figures for the exact machine are still missing from the quotation.
Those are the questions a kitchen can answer for itself in an afternoon, with a notebook and a camera, before any supplier is contacted. Answer them first and the equipment conversation becomes short and specific: you will already know what the floor needs, what it cannot tolerate, and what you will accept as finished.
