Write the cleaning requirement down before the specification sheet
Model selection fails when the requirement stays in somebody's head. Put it on one page: the surfaces in scope, the deposit actually sitting on them, the condition that counts as clean, the area to be covered in one run, the time allowed, and what must not be affected nearby.
Write the run-off question into the same page, because every wet job produces it: where the water goes, what intercepts the soil it carries, and whether that route is lawful for your site. A machine can pass on cleaning and fail on run-off.
This sheet, not the datasheet, is the thing you assess the model against. Everything below is a way of testing one model line against it.
The requirement decides mode before it decides model
Most machines in this category switch between modes, and the published catalogues in this range name dry steam, wet steam and hot water, sometimes as separate modes and sometimes as a dry and wet switch. The deposit chooses between them. Dry steam carries little free water, so it loosens a film and leaves the surface almost workable straight away. Wet steam and hot water put more volume and heat against a thick deposit, and they leave run-off behind.
So the selection question is not which model is strongest. It is which mode carries your duty, and whether the model keeps that mode available for a full job without chasing a refill or waiting for a heater.
Read the specification on the machine being quoted, never on the range
Range pages and variant pages in the same family do not carry the same rows. The Fixed Steam Cleaning Machine PowerDe HW02 page publishes 6 kW heating power, an 80 L SUS304 tank, 220 V working voltage, a footprint of 1,200 x 800 x 1,200 mm including handles, and a tap-water inlet with a four-point internal-tooth joint, 1/2 inch. The hand-pushed units in the same category print a smaller tank and a battery working supply. A figure read on one page does not migrate to another.
Some rows also appear in an unusual form. The HW02 page prints its steam-temperature figure as a bare number with no unit symbol, and its pressure row in a non-standard spelling. Do not tidy that up in your own file; ask for the same value restated in your units, in writing, for the exact machine quoted.
For other models in the range the gaps are larger: no steam temperature, pressure, heater rating, tank volume, battery capacity or runtime published. That is not a fault in the machine, it is a gap in the paperwork, and model selection has to close it by asking rather than by assuming.
Size from observed work, not from the tank row
Tank volume is a container, not a working time. What you need is the number of surfaces, or vehicles, or bays completed on one fill or one charge, measured once with a stopwatch. Compare that figure against the area your requirement sheet names.
Then check what sits next to the capacity row. A machine plumbed to one point depends on the supply, the fitting it terminates in and the drain that has to take the run-off. A mobile unit depends on gate heights, turning space, floor surface and hose reach. The same cleaning result can be blocked by either arrangement, and both are visible on a site walk long before delivery.
Fixed or mobile is a decision about your building
A fixed installation suits work that stays in one place, where the utilities can be run once and the machine never moves. The mobile units in this range are described for indoor garages and narrow outdoor parking spaces, which suits work that has to follow vehicles or cover several positions.
Judge it on your own layout: how many positions have to be cleaned in a shift, how far the route between them runs, and whether one position can take a permanent supply. If your duty moves, a fixed machine with the better numbers can still be the wrong selection.
Match the protection wording to the exact unit
Catalogue pages in this range list protective functions such as leakage, short-circuit, high-temperature, high-voltage and water-shortage protection, and the item list is not identical on every page. Do not treat a protection list as portable across variants, and do not judge it by how many lines it prints.
Ask two practical questions instead. What does the operator see when a protection trips mid-job, and what is the reset procedure? Which of those functions does your own wiring, water and environment actually depend on? A protection you can reset in seconds is a pause; one that needs a technician is a lost shift.
Run the trial against your requirement sheet, not against a demonstration
A demonstration is cleaned by the supplier's operator on the supplier's surface. A trial that informs model selection uses your surface, your deposit, your time allowance and your operator, and it is scored afterwards against the sheet you wrote.
Record five things: which mode did the work, how long the surface took, what was consumed, what was left behind, and which surfaces were excluded and why. Agree the stop condition before the trial starts, so nobody argues about the result afterwards.
Use the trial to test your own method as much as their machine. If the job only works with the supplier's operator holding the wand, the selection has to carry training and written method, and customized steam cleaning equipment is a reasonable thing to raise at that point, because it puts your surfaces and your shift pattern into the specification instead of adapting to a catalogue description.
Score the shortlist in writing
Put every candidate through the same rows, and mark each one stated, qualified, or missing: the duty sentence it is meant to perform; the mode that performs it; the figures for the exact machine quoted; the utility and run-off arrangement at the position it will stand; the consumables with part numbers, stated service interval and lead time to site; the training and handover that survives staff turnover.
A missing row is not a reason to reject a machine. It is a cost that has not been priced yet, and the cheapest quotation is often the one with the most blanks.
Close the selection with written answers
Model selection is settled when the requirement sheet, the trial record and the quoted specification agree. Until then, every unanswered row is something your team will solve on the floor with a bucket, a hose or a phone call, at the moment you can least afford it.
The habit worth keeping is a short one: nothing about a machine counts until it is written down for that machine, at your position, on your surface.
